Open System Services Management and Operations Guide Abstract This guide describes how to manage and operate the HP NonStop™ operating system Open System Services (OSS) environment. Product Version OSS Monitor T8622G11, T8622H01 Supported Release Version Updates (RVUs) This guide supports G06.25 and H06.03 and all subsequent G-series and H-series release version updates until otherwise indicated by its replacement publication.
Document History Part Number Product Version 527191-002 OSS Monitor T8622G11, T8622H01 Published July 2005
Open System Services Management and Operations Guide Glossary Index What’s New in This Guide xiii Guide Information xiii New and Changed Information Figures Tables xiii About This Guide xv What This Guide Is About xv Who Should Read This Guide? xv What This Guide Does Not Cover xv What Is in This Guide? xvi Related Reading, Training, and Services Unsupported Utilities xix Acknowledgment xix Notation Conventions xx xvii 1.
2. Operating the OSS Environment (continued) Contents 2.
4. Managing Servers (continued) Contents 4.
4. Managing Servers (continued) Contents 4. Managing Servers (continued) Removing a Server 4-46 Removing an OSS Name Server 4-46 Removing a Network Services Server 4-47 Troubleshooting a Server 4-47 5.
5. Managing Filesets (continued) Contents 5. Managing Filesets (continued) Troubleshooting Filesets 5-37 Managing and Repairing Fileset Catalog Files Upgrading OSS Catalog Files 5-38 Moving and Removing OSS Catalog Files 5-38 5-39 6.
. Managing Security (continued) Contents 8.
11. Managing Problems Contents 11. Managing Problems Problem-Reporting Procedures 11-1 Gathering Version Information About OSS Files 11-1 12.
A. Messages (continued) Contents A. Messages (continued) FSCK Messages A-6 FSCK Consistent-Fileset Messages A-7 FSCK Inconsistency and Error Messages A-7 OSS Message-Queue Server Messages A-27 OSS Monitor Messages A-30 Unnumbered Messages A-31 Numbered Messages A-38 OSS Subsystem Messages A-61 OSSTTY Subsystem Messages A-95 Startup Messages A-96 EMS Messages A-100 B. Manually Setting Up an OSS Environment C.
E. Environment Limits Contents E. Environment Limits Glossary Index Figures Figure 1-1. Figure 1-2. Figure 1-3. Figure 1-4. Figure 1-5. Figure 1-6. Figure 1-7. Figure 2-1. Figure 2-2. Figure 2-3. Figure 2-4. Figure 2-5. Figure 3-1. Figure 3-2. Figure 3-3. Figure 3-4. Figure 4-1. Figure 4-2. Figure 4-3. Figure 4-4. Figure 4-5. Figure 4-6. Figure 5-1. Figure 5-2. Figure 5-3. Figure 5-4. Figure 6-1. Figure 6-2.
Figures (continued) Contents Figures (continued) Figure 6-3. Figure 6-4. Figure 6-5. Figure 6-6. Figure 7-1. Figure 7-2. Figure 7-3. Figure 8-1. Figure 8-2. Figure 9-1. Figure 10-1. Figure 12-1. Figure 12-2. Figure 12-3. Figure 12-4. Figure 12-5. Figure C-1. Figure C-2. Figure C-3. Figure C-4. Figure D-1.
Tables (continued) Contents Tables (continued) Table C-1. Table E-1.
Contents Tables (continued) Open System Services Management and Operations Guide —527191-002 xii
What’s New in This Guide Guide Information Open System Services Management and Operations Guide Abstract This guide describes how to manage and operate the HP NonStop™ operating system Open System Services (OSS) environment. Product Version OSS Monitor T8622G11, T8622H01 Supported Release Version Updates (RVUs) This guide supports G06.25 and H06.03 and all subsequent G-series and H-series release version updates until otherwise indicated by its replacement publication.
What’s New in This Guide New and Changed Information The description of DSAP use for backup was corrected in Creating a pax Backup of OSS Files in the Guardian File System on page 6-20 to resolve Genesis solution 10050408-6471. Adding Commands for User Convenience on page 9-3 was modified to add information about the behavior of the OSS man command when the MANPATH environment variable is not set.
About This Guide The HP NonStop operating system Open System Services (OSS) environment enables users on HP NonStop servers to integrate an operating system similar to the UNIX operating system into their work environment. You manage and operate the OSS environment primarily from the Guardian environment. What This Guide Is About This guide describes how to manage and operate the OSS environment. It describes only what is unique to the OSS environment, its management, and its operation.
What Is in This Guide? About This Guide What Is in This Guide? This guide contains information and procedures for managing and operating the OSS environment. It is divided into the following sections: • • • • • • • • • • • • Section 1, Introducing Open System Services, presents an overview of how to manage and operate the OSS environment. It includes a task table that refers you to the appropriate documentation for specific kinds of tasks.
Related Reading, Training, and Services About This Guide • • • • • Appendix A, Messages, describes OSS EasySetup, CVT, FSCK, OSS messagequeue server, OSS Monitor, and OSS subsystem messages. Appendix B, Manually Setting Up an OSS Environment, summarizes the procedures to configure and start a new OSS environment. Appendix C, OSS Management Utilities, describes the OSSTTY and OSS EasySetup product utilities and files.
Additional Reading About This Guide • • • • • SCF Reference Manual for G-Series RVUs SCF Reference Manual for H-Series RVUs Security Management Guide Software Internationalization Guide The courses NonStop S-Series Configuration and Change Management and NonStop S-Series Production Management Additional Reading Problem reporting often involves using the Event Management Service (EMS) and the Guardian VPROC utility.
Unsupported Utilities About This Guide • • • System Generation Manual for G-Series RVUs Telserv Manual WAN Subsystem Configuration and Management Manual UNIX System Administration • • • Frisch, Æleen. Essential System Administration, Second Edition. O’Reilly & Associates, Inc. Sebastopol: 1995 Nemeth, Evi, Garth Snyder, Scott Seebass, and Trent R. Hein. UNIX® System Administration Handbook, Second Edition. Prentice-Hall. Englewood Cliffs: 1995 Reiss, Levi and Joseph Radin.
Notation Conventions About This Guide Copyright (c) 1989 Mark H. Colburn. All rights reserved. The pax utility, distributed to HP free of charge, is used to support the tar, cpio, and pax user interfaces furnished with the OSS environment and is documented in this guide and the Open System Services Shell and Utilities Reference Manual.
General Syntax Notation About This Guide [ ] Brackets. Brackets enclose optional syntax items. For example: TERM [\system-name.]$terminal-name INT[ERRUPTS] A group of items enclosed in brackets is a list from which you can choose one item or none. The items in the list may be arranged either vertically, with aligned brackets on each side of the list, or horizontally, enclosed in a pair of brackets and separated by vertical lines. For example: FC [ num ] [ -num ] [ text ] K [ X | D ] address { } Braces.
Notation for Messages About This Guide Item Spacing. Spaces shown between items are required unless one of the items is a punctuation symbol such as a parenthesis or a comma. For example: CALL STEPMOM ( process-id ) ; If there is no space between two items, spaces are not permitted. In the following example, there are no spaces permitted between the period and any other items: $process-name.#su-name Line Spacing.
Change Bar Notation About This Guide horizontally, enclosed in a pair of brackets and separated by vertical lines. For example: proc-name trapped [ in SQL | in SQL file system ] { } Braces. A group of items enclosed in braces is a list of all possible items that can be displayed, of which one is actually displayed. The items in the list might be arranged either vertically, with aligned braces on each side of the list, or horizontally, enclosed in a pair of braces and separated by vertical lines.
About This Guide Change Bar Notation Open System Services Management and Operations Guide— 527191-002 xxiv
1 Introducing Open System Services The Open System Services (OSS) environment provides a user and programming interface similar to that of the UNIX operating system. The OSS environment combines the benefits of the UNIX operating system with the features of the HP NonStop operating system. Open System Services differs from the UNIX operating system in that almost all management and operations activities are performed through Guardian environment commands.
Management Tools Introducing Open System Services Figure 1-1. The Operating System Environments OSS Environment OSS Shell and Utilities Guardian Environment Guardian products, subsystems, and command interpreters OSS Application Program Interface Guardian Application Program Interface OSS File System Guardian File System NonStop Operating System VST011.
Management and Operations Tasks Introducing Open System Services Management and Operations Tasks The only OSS management and operations tasks that must occur within the OSS environment are performing backups, configuring printer aliases, and defining default profiles; the rest can occur from the Guardian environment. Table 1-1 lists common OSS management and operations tasks, along with the location of information on performing those tasks.
Management and Operations Tasks Introducing Open System Services Table 1-1.
OSS File System Concepts Introducing Open System Services Table 1-1.
OSS Files Introducing Open System Services • • • OSS pathnames, which have underlying Guardian filenames. The mapping between OSS pathnames and Guardian filenames is known as filename resolution, and it is done by an OSS name server. The /G directory, which contains filesets for local files in the Guardian namespace, and the /E directory, which contains files on other nodes in the network. The /dev directory being used in a special way. Figure 1-2.
Introducing Open System Services The /G Directory The /G Directory The /G directory provides OSS names for Guardian files on your local NonStop S-series or NonStop NS-series node. Each Guardian filename has a corresponding OSS name of the form /G/volume/subvolume/fileID, where volume, subvolume, and fileID are case-insensitive; for example, $SYSTEM.SYS00.CONFLIST becomes /G/system/sys00/conflist. The /G directory itself is reserved for HP use. You cannot put anything in this directory.
The /dev Directory Introducing Open System Services Figure 1-3 shows how the /E directory on your local node allows remote file access through an Expand network. Figure 1-3. Pathname Resolution for Remote File Access Through the Guardian Expand Network Local Node \NODE1 OSS Environment Guardian Environment Remote Node \NODE2 OSS Environment Guardian Environment OSS pathname OSS pathname Guardian filename /E /node2 /usr /donl /test /usr /donl /test \NODE2 .$VOL2 .ZYQ00011 .
Introducing Open System Services Components to Be Managed Components to Be Managed You need to manage the software described in the following subsections: • • • OSS Security on page 1-9 OSS File-System Components on page 1-9 Interprocess Communication Facilities on page 1-11 Guardian administrative applications or OSS applications at your site might also require you to make one or more copies of an OSSTTY server available for redirecting OSS standard files.
OSS File-System Components Introducing Open System Services Figure 1-4.
Introducing Open System Services Interprocess Communication Facilities An OSS file manager process named $ZFMnn runs in each processor (nn indicates the processor number). The OSS file manager starts automatically when the processor starts; if the OSS file manager terminates abnormally, it takes down the processor. Its processor can be shut down without first stopping the process, but you should stop all applications with open OSS files first.
Interprocess Communication Facilities Introducing Open System Services Figure 1-5 on page 1-12 shows these facilities with their approximate equivalents in the Guardian environment. Figure 1-5.
Interprocess Communication Facilities Introducing Open System Services Figure 1-5.
Introducing Open System Services Interprocess Communication Facilities OSS shell commands allow you to manage OSS message queues, OSS shared memory, and OSS semaphores; these commands are discussed further in Section 2, Operating the OSS Environment. OSS Message Queues Message queues are linked lists of messages used by programmers to pass data from one process to another. In the OSS environment, the OSS message-queue server process named $ZMSGQ manages the message queues.
Introducing Open System Services • Interprocess Communication Facilities Multiple transport-provider processes can be used to create separately addressed IP subnetworks within a node When AF_INET sockets are used with NonStop Parallel Library TCP/IP: • • • No transport-agent process or transport-provider process is involved in data routing. A transport-agent process must still be started in each processor to initialize the OSS sockets interface for that processor.
Introducing Open System Services Interprocess Communication Facilities HP provides the OSS sockets local server, $ZPLS, as a transport-provider process for OSS AF_UNIX sockets. You manage the OSS sockets local server through the SCF module of the OSS Monitor, as described in Section 4, Managing Servers. HP provides processes for each of its TCP/IP implementations as transport-provider processes for OSS and Guardian AF_INET and AF_INET6 sockets.
Introducing Open System Services Interprocess Communication Facilities forth by the transport-agent processes. Transport-agent processes do not communicate with each other directly, but through the transport-provider. Note. Whenever any application process sends a message to another process, the message is always sent through the transport-provider processes and through the network, regardless of which processors or nodes they are running on.
Interprocess Communication Facilities Introducing Open System Services Figure 1-6. OSS AF_INET Sockets Servers for NonStop TCP/IP Node A Processor 0 Processor 1 Process 1 Process 2 $ZTA00 $ZTA01 $ZTC0 Node B Processor 0 Processor 1 $ZTC0 $ZTC1 $ZTA00 $ZTA01 Process 3 Process 4 VST012.
Interprocess Communication Facilities Introducing Open System Services TCP6MON, which runs as a process with the name $ZPTMn (where n is its processor number), must be running in a processor that runs an OSS sockets application program. See the TCP/IPv6 Configuration and Management Manual for more information on data flow between sockets applications. AF_UNIX Sockets AF_UNIX sockets allow an application program to use an OSS socket as if it were a named disk file or a named pipe.
Introducing Open System Services Interprocess Communication Facilities Open System Services Management and Operations Guide— 527191-002 1- 20
2 Operating the OSS Environment You operate the Open System Services environment from the Guardian environment. As a system operator, when you are operating the OSS environment, there is little you need to do differently from operating the Guardian environment. This section describes what you do to operate the OSS environment that is unique to the OSS environment.
Operating the OSS Environment Automatic Startup Service provides the most availability for OSS systems and requires the least user intervention if a failure occurs. • • Using the STARTOSS utility if your system was initially configured by using the OSSSETUP utility, and the OSSINFIL file has been properly maintained. See STARTOSS Utility on page C-14 for more information. Using individual commands described under Managing the OSS Subsystem on page 2-6 and in other sections of this guide.
Operating the OSS Environment Possible Ways to Stop the OSS File System Possible Ways to Stop the OSS File System You can stop the OSS file system by unmounting the root fileset, which effectively stops the OSS environment. This step can be performed either by: • • Using the STOPOSS utility. STOPOSS might also stop processes you want to continue running; see the STOPOSS Utility on page C-16 for considerations when using this command.
Operating the OSS Environment Manually Stopping the OSS File System and the OSS Environment however, the most commonly used signal can be sent by entering the OSS shell command: kill PID1 PID2 PID3 ... where PID1, PID2, and PID3 are OSS process IDs displayed by the ps command.
Operating the OSS Environment Manually Stopping the OSS File System and the OSS Environment Sending Warnings to Users Use the OSS shell wall command or an OSS shell script to warn users of the OSS shell about a shutdown. Follow your site’s broadcast message procedures to warn users of Guardian environment processes that might be using OSS files.
Operating the OSS Environment Manually Restarting the OSS File System and the OSS Environment For information about shell scripts, see the Open System Services User’s Guide. For information about the wall command, see the wall(1) reference page either online or in the Open System Services Shell and Utilities Reference Manual. You should also follow your site’s broadcast message procedures to warn users who log in through TACL to use OSS files from Guardian environment processes.
Operating the OSS Environment Starting the OSS Monitor messages recorded using the Event Management Service (EMS) through either of the following: • • The OSS shell logger command Program calls that use the syslog() function and related functions For more information about the logger command, see the logger(1) reference page either online or in the Open System Services Shell and Utilities Reference Manual.
Operating the OSS Environment Starting the OSS Monitor OSSSETUP utility, you can start $ZSMP by using the STARTOSS utility; see STARTOSS Utility on page C-14 for more information. Alternatively, you can perform these actions yourself. Starting the OSS Monitor as a Normal Process 1. Log in as the super ID. 2.
Operating the OSS Environment Starting the OSS Monitor Starting the OSS Monitor as a Persistent Process 1. At a TACL prompt, enter: SCF ASSUME PROCESS $ZZKRN 2. Start the $ZSMP process as a generic process so that the OSS Monitor can communicate with it without a delay during restart. a. At an SCF prompt, enter the following to add $ZSMP to the NonStop Kernel subsystem configuration as a generic process: ADD PROCESS $ZZKRN.#ZSMP, & PROGRAM $SYSTEM.SYSTEM.
Operating the OSS Environment Starting the OSS Monitor 4. At an SCF prompt, to start the OSS Monitor, enter: START PROCESS $ZZKRN.#ZPMON 5. At an SCF prompt, to verify that the OSS Monitor is running, enter: STATUS PROCESS $ZZKRN.#ZPMON 6. If the OSS Monitor process is not running, check the Event Management Service (EMS) log for related event messages. See Appendix A, Messages, for explanation of any console messages received and possible corrective actions.
Starting the OSS Monitor Operating the OSS Environment Table 2-1. Currently Used TACL PARAMs for the OSS Monitor (page 1 of 2) PARAM Description AUTOSTART { AUTO | MANUAL } Specifies whether the automatic startup service should be used for any stopped servers or filesets that are configured to use that service. This PARAM can provide the initial configuration value for the AUTOSTART attribute of the subsystem.
Starting the OSS Monitor Operating the OSS Environment Table 2-1. Currently Used TACL PARAMs for the OSS Monitor (page 2 of 2) PARAM Description OSS^NAMESERVER^TIMEOUT seconds Defines the amount of time that the OSS Monitor waits for a response from an OSS name server. This PARAM can provide the initial configuration value for the IOTIMEOUT attribute of the subsystem.
Operating the OSS Environment Starting the OSS Monitor Monitor can no longer start an OSS name server or the FSCK utility. A valid terminal device must exist when the OSS Monitor starts these programs. One possible solution to the problem is to select a permanent terminal such as an asynchronous terminal or a terminal-simulator process (such as $ZHOME) and then specify that device as the OSS Monitor home terminal.
Operating the OSS Environment Starting the OSS Monitor The SCF command: INFO /OUT $S.#srv.info/ SERVER $ZPMON.* returns descriptions of all the servers in the ZOSSSERV file. The SCF command: STATUS /OUT $S.#file.stat/ FILESET $ZPMON.*user* and the SCF command: STATUS /OUT $S.#file.stat/ FILESET $ZPMON.*USER* both return the status of every fileset in the ZOSSFSET file that contains the string “USER” in its name. Table 2-3.
Stopping the OSS Monitor Operating the OSS Environment 1. Use the SCF ASSUME command to access the OSS Monitor: ASSUME PROCESS $ZPMON 2. Issue the SCF ALIAS command for the desired alias: ALIAS DF STATUS FILESET 3.
Obtaining Information About the OSS Subsystem Operating the OSS Environment For example, if you enter LISTDEV at an SCF prompt, a display similar to that shown in Figure 2-3 would appear.
Operating the OSS Environment Obtaining Information About the OSS Subsystem Listing the Objects Managed by the OSS Monitor When the OSS Monitor is running, you can use the SCF NAMES command to identify the objects managed by the OSS Monitor.
Operating the OSS Environment Changing the OSS Subsystem Configuration Checking the Version of the OSS Monitor When the OSS Monitor is running, you can use the SCF VERSION command to determine the product-version information for the running copy of the OSS Monitor. For example, if you enter the following SCF commands: ASSUME PROCESS $ZPMON VERSION the system displays basic product-version information for the running OSS subsystem.
Operating the OSS Environment Removing the OSS File System FILESET Command on page 12-19 and ALTER SERVER Command on page 12-26. 3. Use the SCF ALTER SUBSYS, AUTOSTART AUTO command to enable the automatic startup service and initiate automatic startup. See ALTER SUBSYS, ALTER MON, and ALTER PROCESS Commands on page 12-32. Removing the OSS File System This action can be performed in one of these ways: • • Using the procedure described in this subsection. Using the OSSREMOV utility.
Operating the OSS Environment Monitoring OSS Processes 12. Remove all previously saved catalogs by entering the following TACL command: RUN $tsvvol.ZOSS.CVT PURGE SERIAL mmmm IN $old_vol.ZX0nnnnn $tsvvol specifies the disk volume of the installation target subvolume (TSV) containing the CVT utility. See Managing and Repairing Fileset Catalog Files on page 5-38 for more information on the CVT utility. mmmm specifies the FSCK serial number of a previous FSCK diagnose or repair operation.
Operating the OSS Environment Monitoring OSS Processes From the OSS Environment Monitoring OSS Processes From the OSS Environment The OSS shell ps command has a -W flag with many options for monitoring processes from the OSS environment.
Operating the OSS Environment Managing OSS Processes Figure 2-5. TACL STATUS, DETAIL Display for an OSS Process . . . System: \NODE1 June 4, 2001 12:09 Pid: 9,326 ($Z0S5) Primary Priority: 160 Wait State: %001 (LREQ) Userid: 255,44 (SUPER.WEBMASTR) Myterm: $ZTN1.#PT74DAB Program File Name: $ROOT.ZYQ00000.Z00116NL Swap File Name: $ROOT.#0 Process Time: 0:0:0.008 Process Creation Time: February 17, 2001 16:53:25.
Operating the OSS Environment Making OSS Application Processes Persistent with the Kernel Subsystem Making OSS Application Processes Persistent with the Kernel Subsystem Just as the OSS Monitor can be made a persistent process using the SCF interface to the NonStop Kernel subsystem, the same SCF interface provides commands to configure, start, and stop OSS application processes that are persistent.
Operating the OSS Environment Making OSS Application Processes Persistent with the Kernel Subsystem >>out redirects standard output to the file out in the current working directory of the OSS process 2>>err redirects standard error to the file err in the current working directory of the OSS process See the sh(1) reference page either online or in the Open System Services Shell and Utilities Reference Manual for more information about OSS shell redirection.
Operating the OSS Environment Making OSS Application Processes Persistent with the Kernel Subsystem consideration means that an OSS persistent process that runs in more than one processor (an OSS persistent process with a valid CPU attribute) must be started from within a script that uses an OSS shell run command with a -name flag of the form: run -name base_name$ZCPU...
Operating the OSS Environment • Making OSS Application Processes Persistent with the Kernel Subsystem When more than 128 characters are required to correctly start an application or script, a script requiring less than 128 characters in its STARTUPMSG attribute can be run to start the command or script that needs to be made persistent.
Operating the OSS Environment Making OSS Application Processes Persistent with the Kernel Subsystem launches the named process $OSS1 that runs the program (-p) /bin/tail as if it were a login shell (-ls); the tail utility uses the file (-f) named log in the home directory of the user with the user ID OSS.APPS. Because of the -ls flag, the tail command inherits any environment variables defined in either /etc/profile or $HOME/.profile for the user ID OSS.APPS.
Operating the OSS Environment Making OSS Application Processes Persistent with the Kernel Subsystem PROGRAM $SYSTEM.SYSTEM.OSH ASSOCPROC $OSS2, STARTUPMSG "-ls -p /bin/sh startmyapp reload" c. Because the configuration uses a STARTMODE of MANUAL, you must enter the following command at an SCF prompt to start the persistent process: START PROCESS $ZZKRN.#OSSAPP These commands configure and start the persistent process object OSSAPP in processor 2, using the named process $OSH2 to launch the OSH utility.
Operating the OSS Environment Making OSS Application Processes Persistent with the Kernel Subsystem To configure and run myapp in all processors as a persistent process that uses the CONFIGURATION value of reload, enter the following commands: a. From a TACL prompt: WHO which produces output that includes the HOMETERM value you need to use, such as: Home terminal: $ZTN0A.#PT4KH30 ... b. Then, at an SCF prompt, enter: ADD PROCESS OSSAPP, NAME $OSH, HOMETERM $ZTN0A.
Operating the OSS Environment Managing OSS Process Scheduling Managing OSS Process Scheduling A frequently used process might not obtain adequate processor time when it runs with a default priority. If that happens, you can assign a nondefault priority to the process. In the Guardian environment, you would write a CMON process to control the scheduling of other processes. In the OSS environment, you can start processes with nondefault priorities using the OSS nice command.
Operating the OSS Environment Managing OSS Process Scheduling Using an OSS Shell Script Unlike a shell alias, the shell script method works for calls to a program from within another shell script. To use an OSS shell script that changes the default priority of a program, follow these steps: Note. Using shell scripts increases system overhead. Be sure that the performance tradeoff is worthwhile before you use this technique. 1.
Operating the OSS Environment Managing OSS Process Scheduling 4. Post a broadcast message to users, using the technique described in Manually Stopping the OSS File System and the OSS Environment on page 2-3. This procedure can be used to alter the scheduling priority of any process, including those released by HP as part of the OSS environment. For example, the following steps decrease the scheduling priority for the c89 compiler by 15: 1.
Operating the OSS Environment Managing OSS Process Processor Use Managing OSS Process Processor Use You can also use OSS shell scripts or aliases to force specific processes to execute in specific processors to distribute the work load within your NonStop S-series or NonStop NS-series node. Note. Using shell scripts increases system overhead. Be sure that the performance tradeoff is worthwhile before you use this technique. To change the default processor for a specific process: 1.
Operating the OSS Environment Managing OSS Interprocess Communication Facilities 3. Change the PATH environment variable in /etc/profile to something similar to the following: export PATH=/usr/local/script:/bin:/bin/unsupported: /usr/ucb:/usr/bin 4. Post a broadcast message to users. Managing OSS Interprocess Communication Facilities The OSS interprocess communication (IPC) facilities require little management.
Using the cron Process Operating the OSS Environment product with the TACL OSH command. The following subsections discuss these options: • • Using the cron Process on page 2-35 Using the NetBatch Product on page 2-38 Using the cron Process The cron process runs shell commands at specified dates and times. The commands that are to be run and the schedule are placed in entries within files in a specific directory. Commands that are to run only once are placed within files in the at queue.
Operating the OSS Environment Using the cron Process /var/adm/cron/cron.deny Lists the user IDs that are denied access to the crontab file. /var/adm/cron/.proto Contains shell commands required to provide the correct shell environment for at and batch jobs. /var/adm/cron/queuedefs Configures the task queues to be used for the at, batch, and cron commands. HP provides sample files for the at.deny, cron.deny, .proto, and queuedefs files.
Operating the OSS Environment Using the cron Process For more information about the crontab command, see the crontab(1) reference page either online or in the Open System Services Shell and Utilities Reference Manual. The at Command The at command runs OSS shell commands at a time you specify. You can use the at command to read from the standard input file or accept as arguments the names of commands to be run and when the commands are to be run.
Operating the OSS Environment Using the NetBatch Product specify one or more user names, only jobs belonging to those users are displayed. If you do not specify any user names, a list of all jobs submitted is displayed. The atq command allows you to sort the output in chronological order based on the time that the at command was issued. You can also specify which queue you want to have printed.
Operating the OSS Environment • Using the NetBatch Product The general form of the TACL OSH command for batch execution of a program in the OSS environment is: OSH <- >>out_file 2>>err_file -p program_path program_args > indicates that the OSS shell appends normal output to the identified file instead of the standard output (stdout) file. out_file is the OSS pathname of the file to receive normal output.
Operating the OSS Environment Using the NetBatch Product specified, the specified pathname must be a Guardian file identifier and the program file must reside in that subvolume. program_args specifies any arguments expected on a command line by the program_path program. Such arguments would include the name of an input file.
Operating the OSS Environment Using the NetBatch Product If the initial working directory is not the current working directory, relative pathnames are resolved from the initial working directory. When the initial working directory is a Guardian subvolume and only a relative pathname is specified, the specified pathname must be a Guardian file identifier and exception output is saved in a file-code-180 file in that subvolume. Site-written shell programs can also be used instead of the default OSS shell.
Operating the OSS Environment Using the NetBatch Product directory is a Guardian subvolume and only a relative pathname is specified, the specified pathname must be a Guardian file identifier and exception output is saved in a file-code-180 file in that subvolume. script_path specifies the OSS pathname for the OSS shell script file to be run. If the initial working directory is not the current working directory, a relative pathname is resolved from the initial working directory.
Operating the OSS Environment Using the NetBatch Product 2. Enter the following at a TACL prompt: BATCHCOM;SUBMIT JOB, IN \NODE.$SYSTEM.SYSTEM.HOURLY, & EVERY 60 * * * * mon-fri As an example of running an OSS shell script periodically, suppose: • • • You want to record OSS process activity every 10 minutes only on weekdays. You have created an OSS shell script file containing the ps command at /script/processes.sh. You want to record normal output in the OSS file /script/processes.
Operating the OSS Environment Using the NetBatch Product Open System Services Management and Operations Guide— 527191-002 2- 44
3 Understanding the OSS File System To manage the Open System Services (OSS) environment effectively, you must understand the OSS file system. The OSS file system works in the same way as a UNIX file system from the point of view of the user. Files in the OSS environment are organized in a hierarchical tree structure. For further information about using the tree structure, see the Open System Services User’s Guide.
OSS Pathnames Understanding the OSS File System OSS pathnames are logical names; they have no connection to storage devices. The relationship of OSS files to disk volumes is illustrated in Figure 3-1. Figure 3-1. OSS Files and Disk Volumes 1 2 3 5 5 5 2 /sdir1 /sdir1/a /sdir1/a/file1 / (root) 4 ... 4 3 ...
OSS Pathnames Understanding the OSS File System Figure 3-2. Guardian Files and Disk Volumes 1 \MYNODE $OTHVOL 2 3 2 $OURVOL 3 SUBVOL1 4 AFILE1 AFILE2 MYSUBVOL 4 MYFILE1 MYFILE2 \MYNODE.$OURVOL.MYSUBVOL.MYFILE1 3 3 SUBVOL2 4 BFILE1 BFILE2 \MYNODE.$OURVOL.YRSUBVOL.YRFILE1 YRSUBVOL 4 YRFILE1 YRFILE2 Legend 1 2 HP NonStop server node Disk volumes 3 Subvolumes 4 Files VST003.VSD Each pathname for an OSS regular file has an underlying Guardian filename.
OSS Pathnames Understanding the OSS File System define or manage these filesets through the SCF interface and they do not appear in the fileset configuration database. The OSS pathnames of Guardian files can have no more than four elements including the /G, as illustrated in Figure 3-3. These elements, although technically OSS filenames, are subject to the length restrictions for Guardian filenames. Figure 3-3. Guardian Files in the OSS File System / (root) directory /E /G 2 1 ... ... ... ...
Understanding the OSS File System Using Pathnames for Remote Files Using Pathnames for Remote Files The rules described in the preceding subsection can be extended to the files accessible through the /E directory. A file on a remote node that is connected to your node through the Expand network appears in /E when all of the following are true: • • • The remote Expand node has a TOSVERSION of D40 or later. The remote node has an OSS name server running with a device subtype of 5.
OSS File Components Understanding the OSS File System Such a search would be extremely time-consuming and would return undesired information on /E and /G files. OSS shell commands that perform recursive operations include: chgrp chmod chown cp diff find ls mv pax rm rmdir Other products running in the OSS environment might also have commands with recursive behavior.
Understanding the OSS File System • OSS Catalog Files The data portion, which is stored in a file that is identified by a Guardian filename. This portion contains the data seen by end users of the file. The Guardian filename is mapped from the OSS pathname using an OSS name server catalog so that the NonStop operating system disk process can find the data file. Directories, terminal device files, AF_UNIX sockets, and FIFOs do not have underlying data files.
Understanding the OSS File System Relating OSS Files, Filesets, and Disk Volumes For information on accessing files in the /G directory, see the Open System Services User’s Guide. Relating OSS Files, Filesets, and Disk Volumes You perform OSS file-system configuration and administration tasks through the Guardian environment.
Understanding the OSS File System OSS Configuration Files Within that catalog volume, the OSS name server for that fileset uses a Guardian subvolume whose name begins with ZX0. This name is a reserved subvolume name used only by an OSS name server. In this subvolume, the OSS name server for the fileset accesses (and creates if necessary) the catalog files PXINODE, PXLINK, and PXLOG.
OSS Configuration Files Understanding the OSS File System For this reason, a storage-pool file cannot contain more than 20 active volume names (the creation pool) but the volume list maintained by an OSS name server for the fileset (the entire storage pool) can contain up to 256 disk volumes. See FSCK Log File on page 5-24 for more information about the volume list. You should not set up a fileset to use a disk volume that is not always in the storagepool file for that fileset.
Understanding the OSS File System OSS Configuration Files create new files, it uses the new disk volumes as well as remaining space on the other specified volumes. A sample storage-pool file is shown under Creating a Storage Pool on page 5-6.
Understanding the OSS File System OSS Configuration Files Open System Services Management and Operations Guide— 527191-002 3- 12
4 Managing Servers This section describes how to manage the servers used to provide the Open System Services (OSS) environment. Not all servers are managed through the same interface; read the subsections Introducing the OSS Servers on page 4-1 and Configuration Files on page 4-6 before attempting any operation on an OSS server.
The OSS Name Servers Managing Servers Servers in other subsystems use OSS name servers for OSS pathname resolution. Such servers and subsystems include the following: • • • • The iTP WebServer httpd process The Network File System (NFS) HP NonStop TS/MP The SQLCAT process used with SQL/MP Communication with such servers can be affected by the configuration of OSS servers and of the OSS subsystem.
The OSS Sockets Local Server Managing Servers Figure 4-1.
The OSS Transport Agent Servers Managing Servers The backup server process preserves socket access if the primary server process fails. You can control the processor in which the backup server process runs. You can start and stop the OSS sockets local server using the OSS Monitor SCF commands START SERVER and STOP SERVER (see START SERVER Command on page 12-67 and STOP SERVER Command on page 12-84 for detailed information about these commands).
The Network Services Servers and Tools Managing Servers inetd process allows one process to invoke several others, reducing load on the system. inetd simplifies the interface of a server program that it starts, because it duplicates its socket descriptors for an incoming request as file descriptors 0, 1, and 2 before the server application is processed by an exec function call.
Configuration Files Managing Servers program numbers and can be used to modify the status of RPC servers available from your node. RPCINFO provides a means to monitor and change portmap behavior; RPCINFO is required by the same products that require portmap. For more information about the RPCINFO process, see the rpcinfo(8) reference page in the Open System Services Shell and Utilities Reference Manual. Configuration Files Each OSS server has its own configuration file requirements.
Configuration Files Used for the OSS Name Servers Managing Servers Entries in the ZOSSFSET file must follow these rules: • • The name of a fileset: ° Can consist of 1 to 32 uppercase letters and digits (A through Z and 0 through 9). The first character must be a letter. ° Must be unique. The ZOSSFSET file must contain an entry for the root fileset. An entry for the root fileset is automatically created in the ZOSSFSET file.
Configuration Files Used for the OSS Name Servers Managing Servers This attribute increases the risk of losing file labels if the entire disk-process pair for the volume fails. This option is usually used only for a volume that will contain temporary files. When a default ZOSSFSET file is created by the OSS Monitor on a system that was not previously configured, it contains only an entry for a prototype root fileset with the device label 000000.
Configuration Files Used for the OSS Name Servers Managing Servers Attribute Fileset Fileset Fileset (page 2 of 2) Preconfigured or all defaults accepted $OSS $OSS $OSS Single pool volume specified disk1 disk1 disk1 Two pool volumes specified disk1 disk2 disk2 Three or more pool volumes specified disk1 disk2 disk3 / /home 1 /tmp 1 #ZPNS #ZPNS #ZPNS #ZPNS #ZPNH #ZPNS AUDITENABLED 1 OFF OFF OFF BUFFERED 1 LOG LOG CREATE DESIREDSTATE 2 STOPPED STOPPED STOPPED FSCKCPU
Configuration Files Used for the OSS Name Servers Managing Servers If your system has been upgraded from an RVU preceding G05.00, the OSS Monitor automatically creates an initial ZOSSFSET file containing the information from the older system’s ZPCONFIG and ZPMNTTAB files. If your system is upgraded to a G06.24 or later RVU, the initial MAXINODES value is zero; the OSS Monitor changes that value to a more appropriate one as soon as the fileset is started. If your system is upgraded to a G06.
Configuration Files Used for the OSS Name Servers Managing Servers Figure 4-2. Relationship Among OSS Configuration Files, Processes, and Disk Volumes SCF $SYSTEM $SYSTEM.ZXOSSMON.ZOSSFSET 1 4 $ZPMON $SYSTEM.ZXOSSMON.ZOSSPOOL Disk names 5 $DATA1 $DATA2 Application 3 3 $DATA1.ZYQ00000.Z0000011 $DATA1 $DATA2.ZYQ00000.Z0000435 $ZPNS 2 $DATA2 $ROOTCAT $ROOTCAT.ZX000000.PXLINK $ROOTCAT.ZX000000.PXLOG $ROOTCAT.ZX000000.
Configuration Files Used for the OSS Name Servers Managing Servers If your system was ordered with the OSS environment preconfigured, then the OSS Monitor was started at least once before you received the system. That action created ZOSSFSET and might have embedded the factory-default node number in the ALTFILE attributes. After you change the node number, you must check that the OSS Monitor can open the OSS fileset configuration database the next time the OSS Monitor is started.
Configuration Files Used for the OSS Name Servers Managing Servers When a default ZOSSPARM file is created by the OSS Monitor, it contains entries that have the default values described in ALTER SUBSYS, ALTER MON, and ALTER PROCESS Commands on page 12-32. If your system has been upgraded from a G05.00 or later G-series RVU, the OSS Monitor automatically creates an initial ZOSSPARM file containing the information from the old system’s ZPOSPARM.
Configuration Files Used for the OSS Name Servers Managing Servers The ZOSSSERV File The ZOSSSERV file is an Enscribe file containing the OSS Monitor configuration database that stores information about the characteristics of each OSS name server. Entries in this file can be displayed with the SCF INFO SERVER command. This file must be maintained by the system manager using the SCF ADD SERVER, ALTER SERVER, and DELETE SERVER commands.
Configuration Files Used for the OSS Name Servers Managing Servers Beginning with RVU G06.
Configuration Files Used for the OSS Name Servers Managing Servers If your system has been upgraded from a G-series RVU preceding G05.00, the OSS Monitor automatically creates an initial ZOSSSERV file containing the information from the older system’s ZPCONFIG and ZPMNTTAB files. If your system is upgraded to a G06.17 or later G-series RVU, the OSS Monitor automatically upgrades an existing ZOSSSERV file and creates a backup copy of your original file in $SYSTEM.ZXOSSMON.ZOLDSERV.
Configuration Files Used for the OSS Name Servers Managing Servers The Storage-Pool Files The storage-pool files define the disk volumes that each fileset is currently allowed to use when creating new files. The storage-pool files must be created or edited by the system manager using a Guardian text editor. They must be present and must contain valid data before the OSS Monitor is started.
Configuration Database Files Used for the OSS Message-Queue Server Managing Servers • Comments are allowed; see the sample file shown in Figure 5-3 on page 5-7. Additions to or deletions from a storage-pool file take effect only after the fileset is restarted (remounted). Beginning with RVU G06.10, when HP shipped a system with the OSS environment preconfigured, a storage-pool file named ZINSPOOL was defined for the single predefined fileset and specified the disk volume $OSS. Beginning with RVU G06.
Configuration Database Files Used for the OSS Message-Queue Server Managing Servers Beginning with RVU G06.
Configuration Database Files Used for the OSS Sockets Local Server Managing Servers Attribute Single-Enclosure System Multiple-Enclosure System (page 2 of 2) MAXMQID 32 32 MAXMSG 32 * MAXMQID = 1024 32 * MAXMQID = 1024 MAXWAITTIME 0 0 MSGMSIZE 32000 32000 SERVER #ZPLS #ZPLS TYPE LOCAL LOCAL If your system is upgraded to a G06.17 or later G-series RVU, the OSS Monitor automatically upgrades an existing ZOSSSERV file and creates a backup copy of your original file in $SYSTEM.ZXOSSMON.
Configuration Database Files Used for the OSS Sockets Local Server Managing Servers Beginning with RVU G06.10, when HP shipped a system with the OSS environment preconfigured, the #ZPLS entry matched the default, as follows: Attribute Default Values When ZOSSSERV Is Created by the OSS Monitor Initial Values When ZOSSSERV Is Delivered on a Preconfigured System BACKUPCPU 0 0 CPU 1 1 SERVER #ZPLS #ZPLS TYPE LOCAL LOCAL Beginning with RVU G06.
Configuration Database Files Used for the OSS Transport Agent Servers Managing Servers If your system is upgraded to a G06.17 or later G-series RVU, the OSS Monitor automatically upgrades an existing ZOSSSERV file and creates a backup copy of your original file in $SYSTEM.ZXOSSMON.ZOLDSERV. Figure 4-5 on page 4-22 shows how the OSS sockets local server entries in the ZOSSSERV file correspond to processes, disk files, and disk volumes.
Configuration Files for the Network Services Servers and Tools Managing Servers The OSS Monitor updates the OSS transport agent server entries in the ZOSSSERV file each time the OSS Monitor starts. For example, if processors 0 and 1 of a fourprocessor system are up when the OSS Monitor starts, the ZOSSSERV file contains a record for #ZTA00 and a record for #ZTA01.
Configuration Files for the Network Services Servers and Tools Managing Servers These OSS files contain the same kind of information as is used by Guardian sockets programs in the Guardian environment. Guardian sockets services are configured using the following files: $SYSTEM.ZTCPIP.HOSTS $SYSTEM.ZTCPIP.IPNODES (for NonStop TCP/IPv6 only) $SYSTEM.ZTCPIP.NETWORKS $SYSTEM.ZTCPIP.PROTOCOL $SYSTEM.ZTCPIP.RESCONF $SYSTEM.ZTCPIP.SERVICES HP provides a default version of each of these files.
Configuration Files for the Network Services Servers and Tools Managing Servers portmap The portmap process uses the Guardian files configured for TCP/IP processes running in the Guardian environment: $SYSTEM.ZTCPIP.HOSTS $SYSTEM.ZTCPIP.RESCONF HP provides a default version of each of these files. See the TCP/IP Configuration and Management Manual or the TCP/IPv6 Configuration and Management Manual for more information about these files.
Adding a Server Managing Servers Adding a Server Some kinds of OSS servers can be added to the OSS configuration database. In the current RVU, you can add only an OSS name server for a fileset other than the root fileset. To add an OSS name server to the OSS configuration database: 1. Determine the appropriate configuration for the new server. For help in defining an appropriate configuration, see the rules described under The ZOSSSERV File on page 4-14. 2.
Configuring an OSS Name Server Managing Servers Configuring an OSS Name Server You configure an OSS name server by adding or deleting an entry for it in the Enscribe database ZOSSSERV file. Remember that a fileset cannot be managed by more than one OSS name server; however, an OSS name server can manage more than one fileset. To add a new OSS name server to a configuration, follow the procedure described in Adding a Server on page 4-26.
Configuring the OSS Sockets Local Server Managing Servers 3. To restart the server, enter this OSS Monitor SCF command: START SERVER $ZPMON.#ZMSGQ If you use the STARTOSS utility at your site, you can also configure the OSS message-queue server to be started by that utility. To do so, add the server name to the OSSINFIL file. See OSSINFIL File on page C-19 for more information.
Configuring Network Services Servers, Tools, and Applications Managing Servers • • • rshd on page 4-30 rexecd on page 4-31 OSS Sockets Applications on page 4-31 The RPCINFO program is configured using an EDIT file in the Guardian environment; see Configuration Files for the Network Services Servers and Tools on page 4-23 for more information.
Configuring Network Services Servers, Tools, and Applications Managing Servers portmap The portmap process is configured either by creating NonStop operating system DEFINEs for it in the TACL session used to start it or by passing parameter values to it in the command that starts it. You can change the following by specifying an ADD DEFINE command at a TACL prompt before starting portmap: • The TCP/IP (transport-provider) process used by portmap.
Configuring Network Services Servers, Tools, and Applications Managing Servers See the .rhosts(4) and hosts.equiv(4) reference pages either online or in the Open System Services Shell and Utilities Reference Manual for more information about these files. See the rshd(8) reference page either online or in the Open System Services Shell and Utilities Reference Manual for information about setting up the rshd process.
Configuring Network Services Servers, Tools, and Applications Managing Servers • • You must keep track of the process names or OSS process IDs (PIDs) that OSS sockets applications use so that those processes running in a specific processor (and possibly all processes used by the same application) can be stopped before you stop an OSS transport agent server. You might want to run multiple TCP/IP processes for scalability or load-leveling.
Starting a Server Managing Servers be controlled by the environment variable TCPIP_RESOLVER_ORDER, as described in the environ(5) reference page, either online or in the Open System Services System Calls Reference Manual. This environment variable can be set in the .profile file on a user-by-user basis or in /etc/profile for all shell processes that launch OSS sockets programs.
Starting the OSS Message-Queue Server Managing Servers 1. Make sure that you are a member of the super group (255,nnn). 2. If the OSS name server you want to start is not the OSS name server for the root fileset: a. Use the OSS Monitor SCF INFO FILESET $ZPMON.*, DETAIL command to select the fileset the OSS name server manages that has a mount point closest in the file system directory hierarchy to the / directory. See the INFO FILESET Command on page 12-46 for a description of the output. b.
Starting an OSS Transport Agent Server Managing Servers • • Automatically using the automatic startup service (see the ALTER SERVER Command on page 12-26 for more information about that alternative) Manually using the following procedure To start this server manually when the system is brought up or to restart this server if it fails: 1. Make sure that you are a member of the super group (255,nnn) 2. Enter the OSS Monitor SCF command: START SERVER $ZPMON.
Obtaining Information About a Server Managing Servers • • Specifying the PARAM SOCKET^TRANSPORT^NAME at a TACL prompt before starting your OSS shell Defining the OSS shell environment variable SOCKET_TRANSPORT_NAME before starting inetd. The ability to change transport-provider processes is useful when your node runs several copies of its TCP/IP processes and you want to do load-balancing among them.
Determining Whether a Server Is Running Managing Servers Checking Servers in the Guardian Environment That Are Not Administered Through the OSS Monitor To determine whether such a server is running, you use the basic SCF LISTDEV command. (There is no OSS Monitor command to provide this information.) Here is an example: Suppose you want to determine whether a transport-provider process for OSS AF_INET sockets is running. The transport provider is the $ZTCn server, described under OSS Sockets on page 1-14.
Determining the Current Configuration of a Server Managing Servers You can also check configuration information for many servers that are not administered by the OSS Monitor if they appear in the LISTDEV output. Use the following SCF command: INFO PROCESS process-name, DETAIL where process-name is the name of the server process as it appears in the LISTDEV output.
Determining Usage and Configuration of Network Services Servers Managing Servers The following additional information is displayed when you specify the DETAIL option in the INFO SERVER command: • • • • The user who created the server configuration, when it was created, the user who last modified the configuration, and when it was last modified The automatic startup settings for the server (AUTORESTART, DESIREDSTATE, and PERSISTENCECOUNT attributes) If the server is an OSS name server: ° ° ° The size
Stopping a Server Managing Servers Stopping a Server How and when you stop a server depends on the type of server. • • • • • Stopping a Specific OSS Name Server on page 4-40 Stopping the OSS Message-Queue Server on page 4-41 Stopping the OSS Sockets Local Server on page 4-41 Stopping an OSS Transport Agent Server on page 4-42 Stopping a Network Services Server on page 4-42 Other servers used by OSS applications require separate procedures.
Stopping the OSS Message-Queue Server Managing Servers If the OSS name server becomes unstoppable, do one of the following: • • Repair all filesets involved using the OSS Monitor SCF DIAGNOSE FILESET Command, restart them, then stop them again. Add a new fileset to be managed by the unstoppable OSS name server (see the ADD FILESET Command on page 12-8 for command syntax). Start and stop that fileset; for example, if the new fileset is named DUMMY, enter: START FILESET $ZPMON.DUMMY STOP FILESET $ZPMON.
Stopping an OSS Transport Agent Server Managing Servers 2. Warn your users. You can use a method similar to the one described under Manually Stopping the OSS File System and the OSS Environment on page 2-3. 3. Issue the following OSS Monitor SCF command: STOP SERVER $ZPMON.#ZPLS Stopping an OSS Transport Agent Server If you need to shut down the QIO subsystem, you must first stop the OSS transport agent server. For more information about the QIO subsystem, see the QIO Configuration and Management Manual.
Reconfiguring a Server Managing Servers Reconfiguring a Server How and when you reconfigure a server depends on the type of server. • • • • Reconfiguring an OSS Name Server on page 4-43 Reconfiguring the OSS Message-Queue Server on page 4-44 Reconfiguring the OSS Sockets Local Server on page 4-45 Reconfiguring a Network Services Server on page 4-46 You cannot reconfigure an OSS transport agent server. Other servers used by OSS applications require separate procedures.
Reconfiguring the OSS Message-Queue Server Managing Servers 7. Do one of the following: • If the OSS name server is the server for the root fileset, stop all filesets by entering the following OSS Monitor SCF command: STOP FILESET $ZPMON.* • If the OSS name server is not the server for the root fileset, reassign all filesets it manages to another running OSS name server. Use the OSS Monitor SCF ALTER FILESET Command on each fileset to change the NAMESERVER attribute for the fileset.
Reconfiguring the OSS Sockets Local Server Managing Servers 3. If you changed only one or more of these attributes: AUTORESTART, BACKUPCPUOK, or MAXWAITTIME, you have completed the task. The changes take effect immediately. 4. Otherwise, warn your users. You can use a method similar to the one described under Manually Stopping the OSS File System and the OSS Environment on page 2-3. 5. Use the following OSS Monitor SCF command to stop the server: STOP SERVER $ZPMON.#ZMSGQ 6.
Reconfiguring a Network Services Server Managing Servers Reconfiguring a Network Services Server Network services servers ignore changes to configuration files while they are running. Configuration files such as /etc/inetd.conf can be edited while the servers are running. However, to make configuration changes take effect, you must: 1. Warn your users. You can use a method similar to the one described under Manually Stopping the OSS File System and the OSS Environment on page 2-3. 2.
Removing a Network Services Server Managing Servers See the INFO FILESET Command on page 12-46 for a description of the output. 3. Use the OSS Monitor SCF ALTER FILESET Command on each fileset managed by the OSS name server you want to remove to change its NAMESERVER entry to specify another OSS name server. 4. Warn your users to make sure that all their files in affected filesets are closed and all OSS shell sessions using those filesets are terminated.
Managing Servers Troubleshooting a Server Message-Queue Server Messages on page A-27, OSS Subsystem Messages on page A-61, or in the Operator Messages Manual. 3. When you are unsure of the effect of an action on a server, use the STATUS SERVER command DETAIL option to obtain the last internally reported error information for it. Look up that error in Numbered Messages on page A-38.
5 Managing Filesets This section describes how to perform operations on Open System Services (OSS) filesets.
Creating a Unique Fileset Managing Filesets • To create the directory /home/henrysp from within the Guardian environment, enter the following commands at a TACL prompt: OSH -p /bin/mkdir /home/henrysp OSH -p /bin/chmod 0777 /home/henrysp Note that the OSS shell chmod command gives read, write, and search (execute) permission to all users of OSS files within the mount-point directory.
Creating a Unique Fileset Managing Filesets are constantly created, deleted, and opened is a good candidate for the BUFFERED CREATE option. Although using the BUFFERED CREATE option for a fileset provides better performance than not using it, the following disadvantages exist: ° Fast-create filesets can contain only one disk volume, and the catalog must reside on that volume.
Creating a Unique Fileset Managing Filesets • • • • • Decreases the likelihood of needing fileset recovery by decreasing the value used for MAXDIRTYINODETIME from its default value to 10 seconds Makes the fileset automatically restart when necessary if the automatic startup service is used, and uses the processor (specified as -1) used by the disk process for the fileset to perform any fileset repairs needed Specifies that the OSS name server for the fileset retains the results of nonretryable Network Fi
Creating a Unique Fileset Managing Filesets 4. Start the fileset with the SCF START FILESET command. Figure 5-2 on page 5-5 shows how a fileset is mounted at a mount point when the fileset is started. The figure illustrates the effects of the following command: START FILESET $ZPMON.USER1 Figure 5-2. Starting (Mounting) a Fileset Before mounting the fileset USER1 from the volume $DATA5 onto the mount point /usr/utils/src. After mounting the fileset USER1 onto the mount point /usr/utils/src.
Managing Filesets Creating a Storage Pool Creating a Storage Pool You create a storage pool by creating a storage-pool file. You can use any valid Guardian file identifier for the name of a storage-pool file. However, you should not name a storage-pool file OSSPOOL so that your storage pool definition is not overwritten by the sample file in a reinstallation or an upgrade of the OSS environment. To define the storage-pool file that is associated with a fileset record: 1.
Managing Filesets • • • Starting (Mounting) or Restarting Filesets Automatic Restart of Filesets After OSS Name Server Failure on page 5-9 Automatic Restart of OSS Name Servers After Processor Failure on page 5-9 Potential Problems During Automatic Restart of Filesets on page 5-9 Figure 5-3. Example of a Storage-Pool File == == == == == == == == == File: $SYSTEM.ZXOSSMON.OSSPOOL This file is a sample POOL file. A POOL file defines the disk volumes where OSS files of an OSS FILESET can be created.
Managing Filesets Automatic Restart of Filesets During OSS Monitor Startup Automatic Restart of Filesets During OSS Monitor Startup The OSS Monitor uses the ZOSSFSET file to record the state of each fileset (for example, STARTED, STOPPED, or DIAGNOSING). Whenever a fileset is started, an entry is made indicating that the fileset is in the STARTED state. When the fileset is stopped, the state of the fileset is changed to STOPPED.
Managing Filesets Automatic Restart of Filesets by the Automatic Startup Service Automatic Restart of Filesets by the Automatic Startup Service A fileset can be configured so that the OSS Monitor automatically starts that fileset after a system load, regardless of whether the fileset was in the STARTED state. Restarted filesets are automatically repaired if necessary. The automatic startup service can also restart the fileset a maximum number of times during a 10-minute period.
Managing Filesets Potential Problems During Automatic Restart of Filesets FILESET command to determine which filesets remain in the UNKNOWN state after the failure of an automatic fileset restart sequence. The FSCK integrity checker also might fail during a restart. For example, if the FSCKCPU value for the OSS Monitor specifies a different processor than the processor that the OSS Monitor is running on, the specified processor could have failed.
Managing Filesets Auditing a Fileset 1. Verify that at least one of the OSS name server processors is running. 2. Stop the OSS Monitor and restart it. The OSS Monitor invokes the automatic fileset restart sequence again, and the restart should succeed this time. Auditing a Fileset An important component of a secure file system is the ability to trace the history of security-related operations on objects in the system.
Managing Filesets Audited SCF Operations Audited SCF Operations The following SCF fileset operations are audited: SCF Commands Used Actions taken START FILESET and STOP FILESET When an audited fileset is started or stopped, the OSS Monitor generates a mount/unmount record. The mount point pathname is present only in the record generated by use of the START FILESET command.
Managing Filesets Stopping (Unmounting) a Fileset Stopping (Unmounting) a Fileset You stop an existing, mounted fileset to make it unavailable to users. This action is also known as unmounting a fileset. Note. A fileset cannot be stopped until every fileset mounted on it is stopped. You stop a fileset under any of the following conditions: • • • • Before diagnosing the fileset. To make the files in the fileset unavailable to users.
Managing Filesets • Changing the Operating Parameters of a Fileset Changing the Physical Makeup of a Fileset on page 5-19 Changing the Operating Parameters of a Fileset You change a fileset configuration when you enter new values for: • • • • • • • • • • • • Automatic restart (DESIREDSTATE attribute) Fileset recovery utility processor choice (FSCKCPU attribute) Fileset recovery utility log file destination (REPORT attribute) Fileset auditing (AUDITENABLED attribute) Fileset buffering (the BUFFERED att
Managing Filesets Changing the Operating Parameters of a Fileset Changing Fileset Buffering The value used for the BUFFERED option can have a dramatic effect on application performance when there are a lot of calls to OSS file system functions that involve pathname resolution (creat(), unlink(), rename(), chmod(), chown(), mkdir(), and so on).
Managing Filesets Changing OSS File Caching for the Disks of a Fileset Changing the OSS File System Mount Point Changing the mount point can affect the behavior of programs that use OSS files, particularly programs that use the symbolic-link feature. You should notify users well in advance of changing an existing mount point.
Managing Filesets Changing OSS File Caching for the Disks of a Fileset Similarly, to enable OSS file caching for a specific disk volume: 1. At a TACL prompt, enter: SCF 2. At the SCF prompt, enter the following SCF command to stop each fileset involved: STOP FILESET $ZPMON.filesetname Enter this command beginning with the last fileset mounted on the affected fileset. Stop the filesets in the reverse order in which they were last started. Stop the affected fileset last.
Managing Filesets Changing OSS File Caching for the Disks of a Fileset OSS file caching enabled on one disk volume in a given fileset but disabled for another disk volume in that fileset, you cannot predict whether a particular file might be cached. If you want to add disks to a storage pool for a fileset that has OSS file caching disabled: 1. Use the Subsystem Control Facility (SCF) storage subsystem to add the disks to the system. 2. Modify the storage-pool file for the fileset. 3.
Managing Filesets Changing the Physical Makeup of a Fileset Table 5-1.
Managing Filesets Changing the Physical Makeup of a Fileset Changing the Fileset Catalog To assign a new fileset catalog to a fileset: 1. Stop the fileset with the SCF STOP FILESET command, as described in Stopping (Unmounting) a Fileset on page 5-13. 2. Use the SCF INFO FILESET Command with the DETAIL option to determine whether the fileset uses the BUFFERED CREATE option.
Managing Filesets Changing the Physical Makeup of a Fileset Removing a Disk Volume From a Storage-Pool File You remove a disk volume from a storage-pool file when the disk volume becomes too full to safely accommodate new files. Removing the disk volume from the storage-pool file leaves it in the storage pool for the fileset and leaves its existing files available for use. To remove a disk volume from a storage-pool file: 1.
Managing Filesets Checking and Repairing Fileset Integrity 7. Delete all the files beneath the mount point with the OSS shell rm -r command. This step: • • Prevents the OSS file system from using the old catalog entries when one of these files is accessed. Empties the corresponding ZYQ subvolume so that subsequent use of the SCF DIAGNOSE FILESET command does not inappropriately restore the files to the lost+found directory of the fileset. 8.
Managing Filesets When Do You Need to Check Fileset Integrity? The DIAGNOSE FILESET command corresponds to the fsck command of many UNIX systems. Only super-group users (255,nnn) can use the DIAGNOSE FILESET command. When Do You Need to Check Fileset Integrity? You diagnose a fileset to make sure that you can read and write files in it. The Guardian FSCK utility looks at every regular file to verify that it has both a directory entry and data.
Managing Filesets FSCK Log File If the FSCK utility fails, the fileset is put into the UNKNOWN state instead of the STOPPED state. If the DIAGNOSE FILESET command is issued with the OPTION STOP option and the fileset being diagnosed has a problem that has not been corrected, a subsequent mount of the fileset might fail. FSCK Log File The FSCK utility writes its output to a Guardian log file. Figure 5-4 on page 5-25 shows two examples of FSCK log files.
Managing Filesets FSCK Log File Figure 5-4. FSCK Log File Examples (page 1 of 2) FSCK - T8621G09 - (15DEC2001) OSS Fileset Validate/Repair Utility Copyright Compaq Computer Corporation 1994, 1995, 1996, 2001 August 31,2001 01:14:53 Options Selected are: CATALOG $OSS.ZX000000, REPAIR ALL, OPEN FSCK Serial Number = 0005 Catalog Subvolume is \NODE1.$OSS.
Managing Filesets FSCK Log File Figure 5-4. FSCK Log File Examples (page 2 of 2) FSCK - T8621G09 - (15DEC2001) OSS Fileset Validate/Repair Utility Copyright Compaq Computer Corporation 1994, 1995, 1996, 2001 August 31,2001 01:14:53 Options Selected are: CATALOG $OSS.ZX000000, STATUS, ROOT Catalog Subvolume is \NODE1.$OSS.
Managing Filesets Inconsistencies Checked by FSCK superblock or updated when FSCK is run. The volume list is used by the OSS name server for that fileset to refresh information about file opens within that fileset.
Managing Filesets Inconsistencies Checked by FSCK Table 5-2. Inconsistencies Checked by FSCK (page 1 of 4) Inconsistency Type Explanation Missing or corrupt superblock Serious If the superblock (node #1) is missing from the PXINODE file, the fileset is unusable. A new superblock must be constructed. Nonzero mounted flag Minor FSCK creates the new superblock; you do not need to do anything. Each time a fileset is mounted for read/write access, its mounted flag is set to a nonzero value.
Managing Filesets Inconsistencies Checked by FSCK Table 5-2. Inconsistencies Checked by FSCK (page 2 of 4) Inconsistency Type Explanation Missing inode Serious There are references to an inode but the inode itself does not exist. FSCK corrects this inconsistency as follows: • • • • • • If the missing inode is /, /E, /G, /lost+found, or /dev, it is added as a directory. If the missing inode is /dev/tty, that special file is added. If the missing inode is /dev/null, that special file is added.
Managing Filesets Inconsistencies Checked by FSCK Table 5-2. Inconsistencies Checked by FSCK (page 3 of 4) Inconsistency Type Orphan ZYQ file (continued) Explanation If the number of links is zero and the file is not currently open, FSCK purges the file. If the link count is nonzero, FSCK does the following: 1. An inode is allocated for the file. 2. The disk process copy of the link count is set to one. 3. The file is placed in /lost+found under a synthesized OSS filename.
Managing Filesets Generated Catalog Files Table 5-2. Inconsistencies Checked by FSCK (page 4 of 4) Inconsistency Type Explanation Invalid inode number Serious or minor A record in the PXINODE file has an inode number that is one of the following cases: • • • Less than or equal to 0 (zero) Greater than or equal to 231 In the reserved inode range 8 through 31 If the inode number is <=0 or >= 231, the inode is discarded along with any links to the inode. These cases are serious inconsistencies.
Managing Filesets What Happens When Diagnosis Appears to Fail? For information about the CVT utility, see Managing and Repairing Fileset Catalog Files on page 5-38. What Happens When Diagnosis Appears to Fail? Errors occurring on ZYQ files cause FSCK to issue a warning and continue. In these cases, the new catalog might still contain minor inconsistencies. Any other errors occurring during writes to the new catalog files cause FSCK to issue an error message and terminate.
Managing Filesets Renaming a Fileset 2. If the fileset is mounted, stop the fileset by using the procedure described under Stopping (Unmounting) a Fileset on page 5-13. 3. Determine the device identifier of the fileset with the SCF INFO FILESET Command. 4. Determine whether you have a current backup of the files in the fileset. If not, create a backup of those files using a procedure described under Backing Up and Restoring OSS Files on page 6-11. 5. Delete the fileset with the SCF DELETE FILESET Command.
Managing Filesets Removing Older Configuration Files Table 5-3.
Managing Filesets Moving a Directory Hierarchy to Its Own Fileset The following example shows how to move the /home directory from the ROOT fileset to a newly-defined HOME fileset. To move the /home directory, you need to be logged in as the super ID. 1. Using the OSS shell command mv, change the name of the /home directory to a file or directory name that does not already exist in the ROOT directory, for example, /homex.
Managing Filesets Cleaning Up a Fileset You can verify the results of the copy by using the following OSS shell commands: cd /homex ls -lR * > /tmp/homex.list cd /home ls -lR * > /tmp/home.list diff /tmp/home.list /tmp/homex.list If the copy was executed correctly, the diff command produces no output, meaning that the two directories are identical. 7. Remove the /homex directory and all files and directories underneath it: rm -r /homex 8.
Managing Filesets Troubleshooting Filesets b. Delete all the files in the ROOT fileset by using the FSCK utility: FSCK PURGE 6. Go back to SCF and start the ROOT fileset again: START FILESET $ZPMON.ROOT 7. Restore the ROOT fileset from the backup. If the pax backup was done to the Guardian file system, you can restore the ROOT fileset directly. See Restoring a pax Archive of OSS Files Directly From the Guardian Environment on page 6-25 for more information.
Managing Filesets Managing and Repairing Fileset Catalog Files For example, the following messages result from the START FILESET example in Step 1 when the OSS Monitor attempts to start the new fileset but must back out the operation because the command fails: 2001-07-26 13:41:50 \NODE1.$ZPMON TANDEM.OSS.D40 -00003 $A , STATE changed from Stopped to Started because of Operator Request. 2001-07-26 13:41:50 \NODE1.$ZPMON TANDEM.OSS.
Managing Filesets Moving and Removing OSS Catalog Files current RVU but not available in the RVU used at the time the fileset was created or last upgraded. For example, a fileset created on a system running a D3x RVU cannot contain symbolic links until its catalog file is upgraded. Your fileset catalog files should have been automatically upgraded when you installed a new product version of the OSS Monitor and started it for the first time.
Managing Filesets Moving and Removing OSS Catalog Files CVT also enables you to: • • • • • Move catalog files between subvolumes on the same disk volume Move catalog files onto and off of ZX0 subvolumes Save a current catalog in the same manner that FSCK does when it saves an existing catalog and creates a new one Make a saved catalog the current catalog for a fileset Purge catalog files saved by FSCK CVT is an unlicensed privileged program and, therefore, can be run only by the super ID (which is 255,
Managing Filesets Moving and Removing OSS Catalog Files RENAME files1 [ TO ] files2 is the RENAME command, which moves (renames) catalog files between and within subvolumes. files1 specifies the files you want to move (rename). files2 specifies the new locations (names) for the files being moved (renamed).
Managing Filesets Moving and Removing OSS Catalog Files Using the CVT HELP Command To get help information for CVT, enter one of the following commands at a TACL prompt: RUN $vol.ZOSS.CVT RUN $vol.ZOSS.CVT HELP RUN $vol.ZOSS.CVT ? $vol is the name of the disk volume where the ZOSS subvolume is located. The CVT HELP command produces the following output.
Managing Filesets • Moving and Removing OSS Catalog Files To restore from a backup tape a catalog that has been destroyed due to a double media failure: 1. Use the Guardian RESTORE utility VOL keyword to restore the catalog to a subvolume whose name does not begin with ZX0. 2. Move the catalog onto the ZX0 subvolume with a TACL command of the following form: RUN $SYSTEM.ZOSS.CVT RENAME CURRENT IN nonZX0subvol TO CURRENT IN ZX0subvol & 3.
Managing Filesets Moving and Removing OSS Catalog Files Open System Services Management and Operations Guide— 527191-002 5- 44
6 Managing OSS Files To manage files in the OSS file system, you need to: • • • • • • • Obtain specific information about their size, location, and fileset membership, as described in Obtaining Information About OSS Files. Install new or updated product files, as described in Installing New Product Files on page 6-4. Remove obsolete files, as described in Removing Obsolete OSS Files and Directories on page 6-9.
Managing OSS Files Interpreting Guardian Filenames for OSS Files The OSS shell gname and pname utilities allow you to provide a Guardian filename and obtain the OSS pathname for a specific file and vice versa. See the following subsections for more information: • • Using the OSS gname Command on page 6-2 Using the OSS pname Command on page 6-3 You can also use the Guardian VPROC utility or the OSS shell vproc command to determine the product-version information for an OSS file supplied by HP.
Managing OSS Files Using the OSS pname Command Figure 6-1. OSS gname Command Examples $ gname test gname: test --> \NODE1.$VOL.ZYQ00000.Z0000DV3 $ gname tes* gname: test2 --> \NODE1.$VOL.ZYQ00000.Z0000KHP $ gname -s test \NODE1.$VOL.ZYQ00000.Z0000DV3 $ gname /E/node2/usr/test3 gname: /E/node2/usr/test3 --> \NODE2.$DATA.ZYQ00001.Z0000DV2 When more than one OSS pathname matches the possible wildcard expansion, only the last file with a matching pathname is listed.
Managing OSS Files Installing New Product Files display and next to the SECURITY heading in the FUP INFO, DETAIL display. For information about interpreting the OSS permissions, see the Open System Services User’s Guide. Figure 6-3. FUP INFO Displays for OSS Files 3> fup info $VOL.ZYQ00000.Z0000DV4 CODE EOF LAST MODIF $VOL.ZYQ00000 Z0000DV4 OSS 142 10:55 OWNER RWEP TYPE REC BL 254,254 -rw-rw-rw- 4> fup info $VOL.ZYQ00000.Z0000DV4, detail $VOL.ZYQ00000.
Managing OSS Files Installing New Product Files This set of archive files is large—on the order of tens of megabytes. It is, however, a temporary group of files. Unless DSM/SCM maintains the OSS files on your node, this set of files can be deleted after OSS installation is complete; deleting them when DSM/SCM maintains OSS files can slow down subsequent installations or upgrades because DSM/SCM will replace them.
Managing OSS Files Using COPYOSS Table 6-1.
Managing OSS Files Using COPYOSS From the archive files, COPYOSS copies the OSS product files into the OSS file system, placing them in locations where you would find them on a typical UNIX system—for example, into directories such as /bin, /usr/ucb, and /usr/include. This command completely installs the basic OSS product set and all other products that use the same installation subvolume. • To load only the Java servlet product files, enter these commands from a TACL prompt: $tsvvol.ZOSSUTL.
Managing OSS Files Using PINSTALL archive file. You can use the -cvf flags of the PINSTALL command to display the table of contents for the pax archive to determine if the named file is actually an empty directory.
Managing OSS Files Removing Obsolete OSS Files and Directories and should be ignored. Using PINSTALL on such files would overwrite current files with obsolete ones. • • • PINSTALL is installed as a part of the basic OSS product set and is available even if the subvolume containing COPYOSS has been deleted from your system.
Managing OSS Files • Updating the whatis Database Files DSM/SCM does not install and maintain OSS product files in your OSS file system, you must manually remove obsolete files after installing any new release version, RVU, or software product revision (SPR) and before using the merge_whatis command. You manually remove obsolete files by entering the following OSS shell commands: cd /etc/install_obsolete Pcleanup -r source Note that the command Pcleanup starts with an uppercase letter.
Managing OSS Files Backing Up and Restoring OSS Files You create or update a whatis database file by merging whatis database fragment files using the merge_whatis utility. Each whatis database fragment file has an OSS filename with the form whatis.piece, where piece varies according to the product containing the reference page files.
Managing OSS Files • • Considerations The pax utility does not support labeled tapes and imposes requirements for unlabeled tapes. The pax utility does not support filenames longer than 100 characters in archives using its default USTAR format. The pax utility can be used when any of the following conditions exists: • • • The system that will be restored (target system) might not have the same node name or node number as the system being backed up (source system).
Managing OSS Files Considerations You cannot use pax on labeled tapes. If you need to backup OSS files to labeled tapes or restore OSS files from labeled tapes, use Backup and Restore 2.0. You are also restricted when using pax to single write operations on unlabeled tape.
Managing OSS Files Backing Up the OSS Environment Using a Version of Backup/Restore A single archive can span more than one reel. The pax utility issues this message during reel switching: Mount next tape to continue? Backing Up the OSS Environment Using a Version of Backup/Restore 1. Record all configuration information for the OSS environment.
Managing OSS Files Backing Up User Files INFO ALIAS * LOG Edit these log files to create OBEY files for Safeguard so that the attributes of each file-sharing group, user ID, and user ID alias can be recreated if necessary. ° Using the T9074 Backup and Restore utilities (if the node you are backing up uses the same RVU of Safeguard as the node you will be restoring). The relevant user ID information is kept in the following files: $SYSTEM.SYSTEM.USERID $SYSTEM.SYSTEM.USERIDAK $SYSTEM.SAFE.
Managing OSS Files Backing Up User Files where /user contains all working directories. Note. Do not begin find or pax operations at the root (/) directory. find and pax perform recursive operations within directories unless you use the UTILSGE=NOG:NOE environment variable or the -W NOG and -W NOE flags. The / directory contains /G and /E.
Managing OSS Files Backing Up User Files -v writes archive member pathnames to the standard error file. -f archive_name specifies the pathname of the output archive, overriding the default standard output file. Guardian tape devices can be specified with the /G naming convention (for example, /G/tape). If the -a flag is also specified and a disk archive medium is used, files are appended to the end of the archive.
Managing OSS Files Backing Up User Files Backup and Restore 2.0 writes the following OSS files to the tape identified by the DEFINE =MYTAPE: • All files larger than 200 Kilobytes belonging to the user dev.user2 in the directories: /user/bin /home/sv/myfile /usr/local/bin • All files from the directories /etc/rc and /var/x modified since January 1, 1999. You can also backup entire filesets using the fileset mount points. See the Backup and Restore 2.
Managing OSS Files Backing Up User Files Restoring OSS Files Using Backup and Restore 2.0 You can specify part or all of multiple directory hierarchies and selectively restore files from within them using Backup and Restore 2.0. For example, if you enter the following at a Backup and Restore 2.
Managing OSS Files Backing Up User Files -f archive-name specifies the relative OSS pathname of the archive file to be read. For example, to read the archive a.pax, with all files rooted in /usr within the archive and extracted relative to the current directory, enter the OSS shell command: pax -r -s ',^//*usr//*,,' -f a.
Managing OSS Files Backing Up User Files To determine the size of the HOME fileset, for example: 1. Determine the subvolume name of the fileset by obtaining its DEVICELABEL value from the fileset definition. The last five digits in the Guardian subvolume name (which begins with ZYQ) correspond to the DEVICELABEL value. For the HOME fileset in this example, the DEVICELABEL is 000001, so this fileset’s subvolume name is ZYQ00001. 2.
Managing OSS Files Backing Up OSS Files to Other Expand Nodes In Figure 6-4, the total number of pages for the HOME fileset is shown as 485776. Multiplied by 2048, the total size of the HOME fileset would be 994,869,248 bytes if it is completely contained on this one disk. Because this is smaller than the 2-GB limit, the HOME fileset can be backed up in the Guardian file system. For more information about the DSAP command, see the Guardian Disk and Tape Utilities Reference Manual.
Managing OSS Files • • OSS Files and Backup/Restore Utilities (T9074) The disk volumes used by OSS on the system to be restored have the same names as the disk volumes to be used on the system to be restored. The size of the each disk volume to be restored on the target system is the same as the corresponding disk volume on the source system. In volume mode, BACKUP backs up only what is on the volume being archived.
Managing OSS Files Restoring User Files Restoring User Files The following subsections describe: • • • • • • • Restoring NonStop SQL/MP Programs Using Backup/Restore Utilities (T9074) on page 6-24 Restoring Security Data Used For File Access on page 6-24 Restoring a pax Archive of OSS Files Directly From the Guardian Environment on page 6-25 Restoring a Compressed pax Archive of OSS Files From the Guardian File System on page 6-25 Restoring Files From a pax Archive to the Guardian File System on page 6-
Managing OSS Files Restoring User Files 2. Use the File Utility Program (FUP) to rename the old security database files and to move the restored copies to the correct subvolume. 3. Stop and restart Safeguard and all Expand lines to close the old security database files and open the restored ones.
Managing OSS Files Restoring User Files Guardian files can be specified as pathnames in a /G directory. Files can be restored within /G directories, but existing Guardian files are overwritten only if you use the -W clobber flag. The Guardian file attributes are not preserved. Guardian files are restored as unstructured Guardian files with the file code 0. Only Guardian files with file codes that are supported by the OSS environment are processed.
Managing OSS Files Redirecting OSS Standard Files Verifying a Restored OSS File Backup The dircmp utility reads two directories, compares their contents, and writes the results to the standard output file. Use dircmp to determine whether the contents of two directories differ in any way, such as when you restore backed-up files to a new location and want to be sure the contents are copied correctly. The dircmp utility compares the filenames in each directory.
Managing OSS Files Redirecting OSS Standard Files OSS standard file); to a Guardian process or the Guardian file system, the OSS process appears to be a Guardian file-system object. OSSTTY can be started by an individual user through the OSH command or directly by an OSS administrator. OSSTTY can be used to redirect one or more OSS standard files to Guardian EDIT files or Guardian processes.
Managing OSS Files Controlling the Maximum Number of Files Figure 6-6. Redirecting All OSS Standard Files /G/ZTTY/#stdin a St nd d ar p in u $VHS ile tf IN /G/ZTTY/#stdout OSS Application Program Sta n OSSTTY Standard output file da rd err /G/ZTTY/#stderr or OUT $DATA.OSSLOG. EDITFIL TE RM file $VHS Controlling the Maximum Number of Files The fileset catalog determines the number of regular files and directories (the number of inodes) a fileset can contain.
Managing OSS Files Controlling the Maximum Number of Files FILESET Command is used. A fileset that has a only a few inodes (enforced by using a small MAXINODES value) can be recovered faster than a fileset that has a large number of inodes. To change the approximate maximum number of inodes permissible in a fileset that is not started: 1. Use the SCF INFO FILESET Command with the DETAIL option to determine the configured value for the MAXINODES attribute. 2.
7 Managing Terminal Access This section briefly describes the concepts and methods available for providing terminal users with access to the Open System Services (OSS) environment. • • • How Users Gain Access to the OSS Environment on page 7-1 Configuring Telserv Access on page 7-2 Configuring FTP Access on page 7-5 How Users Gain Access to the OSS Environment A user gains access to the OSS environment through a server process.
Managing Terminal Access • Configuring Telserv Access The way the initial working directory is configured for a user definition that is not anonymous. (The FTP user can use an FTP client quote command to gain access to the opposite environment.) See Section 8, Managing Security, for information about configuring initial working directories. Configuring Telserv Access Configuring Telserv access is fully described in the Telserv Manual.
Managing Terminal Access Configuring a Telserv Direct Service Configuring a Telserv Direct Service A Telserv process can be configured to provide login either directly to the OSS shell or directly to the command interpreter code of any executable program accessible through the OSS file system. Logging In to the Shell Directly The services feature of Telserv provides OSS users with the ability to log in directly to the OSS shell, bypassing the TACL prompt.
Managing Terminal Access Configuring a Telserv Direct Service Use this option if you do not want availability of the service known to all users. • • HP recommends that you do not configure resilient windows for OSS logins. The default configuration for a service is to not have resilient windows, so simply omit the RESILIENT option when defining the service. Do not rely on the default value for the OSS priority in the SCF ADD SERVICE command; instead, set the priority to a value below 160.
Managing Terminal Access Configuring FTP Access instead of the INITIAL-PROGRAM value from either a third-party product or Safeguard user definition, as described in How Users Gain Access to the OSS Environment on page 8-10. The default value used by the OSH program when the PARAM option does not appear places the user in the OSS shell (/bin/sh). Therefore, you must explicitly define a value for the PARAM option to start any program other than the OSS shell.
Managing Terminal Access Configuring FTP Access OSS FTP client users should have the full OSS pathname (/usr/ucb) of the FTP client program added to the PATH environment variable in their .profile files. The default value for PATH does not include the FTP pathname. Alternatively, you can add /usr/ucb to the /etc/profile file for all users (as is done in the /etc/profile.sample file supplied by HP). Section 8, Managing Security, describes how to set up the /etc/profile file.
8 Managing Security This section covers: • • • • Common and Unique Characteristics of OSS and UNIX Security on page 8-1 Managing Users and Groups on page 8-8 OSS Security Auditing on page 8-23 Protecting Your System on page 8-26 Common and Unique Characteristics of OSS and UNIX Security Basic file security is the same for the OSS environment as on a UNIX system. Files are accessed according to a file mode and access permissions, as described in the Open System Services User’s Guide.
Managing Security Administrative Files and Directories modem ttyda or ttydfa ttys0 through ttys9 These files also can be the target of UNIX security intruders. • The OSS environment does not use the following files and directories sometimes found on UNIX systems: ° In /etc: aliases dfs/dfstab exports ftpusers hosts.lpd mail/aliases or sendmail/aliases rc, rc*, or rc?.d shells syslog.conf system ttys tftpdaccess.cf ° In any of various directories: .plan .project sendmail.
Managing Security Administrative Files and Directories Access to OSS auditing logs occurs through the Safeguard audit reduction tool (SAFEART) program. UNIX directories and files such as the following are not provided and therefore do not require monitoring: ° In /var: aculog adm/acct, adm/lastlog, adm/loginlog, adm/messages, adm/pacct, adm/utmp or adm/utmpx, adm/wwtmp or adm/wtmpx spool/atrun or spool/ftp sulog, vold.log, or xferlog ° In /usr: adm/wtmp (FTP login log) etc/rpc.
Managing Security ° Administrative Tools In the user’s home directory: .netrc .rhosts Take normal security precautions with these files and directories. Administrative Tools On many UNIX systems, certain commands and utilities either exist for security administration or have security considerations. Most such commands and utilities are absent from the OSS environment.
Managing Security Administrative Tools rdist rexd rlogin or rlogind routed or gated sendmail share telnet (a TACL TELNET command exists for the Guardian environment) tftpd tip ulimit usermod /usr/etc/exportfs or /usr/etc/showmount vipw • • The GNU suite of utilities The following environment variables: maxuproc nfs_portmon • • The /bin/sh -r restricted shell option For the inetd program: ° ° -t tracing option Service of systat for port 11 Note that some of these commands and utilities can provide m
Managing Security Users and Groups The OSS shell supports the IFS environment variable but appropriately clears its settings. The OSS inetd program provides the following default services: time echo discard daytime chargen Users and Groups The OSS environment does not provide: • The following predefined or generic user names (accounts): agent bin daemon demo finger ftp games guest help ingres lp mail maint manager news nobody nuucp .
Managing Security Components of OSS Security Management The OSS environment does not provide common UNIX default user names and user IDs unless they are explicitly created by a site administrator. However, equivalent OSS user names and user IDs do exist. For example, the privileges normally associated with the UNIX user name root and the user ID of 0 exist for the OSS user ID (UID) of 65535 (the super ID), which is usually the user SUPER.SUPER.
Managing Security Managing Users and Groups Figure 8-1. Major Components and Interfaces for OSS Security Management OSS Environment Application Using OSS APIs Guardian Environment Access control and information retrieval Manager interface Security Database File auditing $ZSMP File I/O SAFECOM File I/O System Audit Log OSS Regular File SAFEART Guardian File VST 020.VSD See the Security Management Guide for an overview of both Guardian and OSS security.
Managing Security • • • • • Differences Between OSS and UNIX User and UserGroup Configuration Differences Between OSS and UNIX User and User-Group Configuration on page 8-9 How Users Gain Access to the OSS Environment on page 8-10 User and User-Group Attributes on page 8-11 Assigning an Initial Working Directory on page 8-13 Assigning an Initial Program on page 8-17 Hints and suggestions for using these concepts are given in Hints and Suggestions on page 8-19.
Managing Security How Users Gain Access to the OSS Environment A NonStop operating system user belongs to a primary group and can belong to more than one file-sharing group. File-sharing groups other than the primary group are called supplementary groups in POSIX terminology, although that term does not appear in Safeguard manuals. All groups configured for the user make up the user’s group list. By default, the primary group for a new user is the administrative group of the user.
Managing Security User and User-Group Attributes Telserv provides access in the following ways: • • • Indirectly, when the user selects the TACL service, logs in to the Guardian environment from a TACL prompt, and then enters the OSH command. (See the osh(1) reference page either online or in the Open System Services Shell and Utilities Reference Manual.) Directly, when the user logs in to a direct service that invokes the OSH command to start an OSS shell.
Managing Security ° ° • User and User-Group Attributes A file-sharing user group has a group number in the range 0 through 65535. A user definition can have a primary user group that is different from its administrative user group. Each user definition has both a user name and a user ID. ° A user name (sometimes called a logon name) has the form group-name.member-name. ° A user definition has a user ID that is unique within the NonStop node.
Managing Security • • Assigning an Initial Working Directory Initial working directory (if none is defined, the default value is null). Some servers do not use this attribute value; instead, they provide alternative attributes. Initial program (if none is defined, the default value is null). Some servers do not use this attribute value; instead, they provide alternative attributes.
Managing Security Assigning an Initial Working Directory 2. Use SAFECOM to add the OSS pathname of the newly created initial working directory to the user definition as the value for the initial working directory attribute. Be careful when assigning an initial working directory. Proofread the assignment for typographical errors and remember to create the initial working directory in the OSS file system.
Managing Security Assigning an Initial Working Directory For example, the following SAFECOM command assigns the initial working directory /home/henrysp to the user definition with the user name SCRIBES.HENRYSP: ALTER USER scribes.henrysp, INITIAL-DIRECTORY /home/henrysp You can assign an initial working directory to an alias instead of to the user definition itself by using the SAFECOM ALTER ALIAS command.
Managing Security Assigning an Initial Working Directory Figure 8-2. TACL Macro to Configure an OSS User (page 1 of 2) ?TACL MACRO == == This unnamed macro adds FTP user logins to the system. == == #FRAME #SET #INLINEPREFIX> INLECHO ON == == SAFECOM /INLINE/ == == == Add a user definition for the user LINDA in user group == SCRIBE. == == This definition provides the user with only Guardian access == and is appropriate for a user who wants initial FTP == access to the Guardian file system.
Managing Security Assigning an Initial Program Figure 8-2.
Managing Security Assigning an Initial Program Assigning an Initial Program Using Safeguard You can assign an initial program to an existing user definition by using the SAFECOM ALTER USER command to change the value of the INITIAL-PROGRAM attribute. You use a SAFECOM command such as the following: ALTER USER group-number,member-number, INITIAL-PROGRAM /code-dir/program or ALTER USER group-name.
Managing Security Hints and Suggestions Assigning an Initial Program Using a TACLCSTM File You can also assign an initial program to a user by either inserting the following entry in the user’s TACLCSTM file or having the user enter the following at a TACL prompt: RUN OSH -ls -prog pathname where pathname is the pathname of an existing OSS program file to be used as the initial program. The -ls specification causes execution behaviors appropriate for a UNIX shell program.
Managing Security Hints and Suggestions This method is preferable to allowing many users to access the super ID logon. The process for licensing the OSS Monitor is: 1. At a TACL prompt, set the volume to $SYSTEM.SYSnn, where nn indicates the current system subvolume. 2. Enter the following commands: FUP LICENSE OSSMON FUP SECURE OSSMON, "N-G-", PROGID This procedure allows anyone in the super group to start OSSMON but also allows OSSMON to run under the super ID.
Managing Security Hints and Suggestions Configuring FTP Users If an FTP server user requests that his or her user definition be configured for initial access to the Guardian environment: 1. Leave that user’s definition with the default null value for its OSS initial working directory. 2. Use the TACLCSTM file method to assign an OSS initial working directory for indirect Telserv access (see Assigning an Initial Working Directory Using a TACLCSTM File on page 8-17).
Managing Security • • Hints and Suggestions An alias must have a valid initial working directory (in the previous example, /user/guest): ° If no valid initial working directory is specified for an anonymous alias, then FTP access for that alias is denied. ° An initial working directory of /E, /G, or a directory in /G is invalid. The aliases anonymous and ftp must be frozen under the Safeguard product, so that those names cannot be used for access through any server process other than FTP.
Managing Security OSS Security Auditing This definition prevents use of the Guardian environment default volume and subvolume as the initial working directory. It also causes use of the OSH command to fail for that user and give the user the suggested pathname as part of an OSH error message. Note that this approach to user and alias definitions can add significantly to logs of error messages or of process failures.
Managing Security • Audit Records for OSS Objects The operation, including whether the operation succeeded or failed, and the details of a defined list of appropriate attributes Audit records are characterized by the following information: • • • • • • • If the object of an operation has a pathname, then either the audit record includes the pathname or the operation is associated with another record that includes the pathname. OSS filenames stored in the audit record are uniquely identifiable.
Managing Security Audit Records for OSS Objects The OSS name server maintains the absolute pathname of the mount point for each fileset that it manages. To ensure that they are generated quickly, all pathnames that are stored in audit records are normalized as follows: • • All dots (.), double dots (..), multiple slashes, and symbolic-link references are resolved. The maximum length of the stored pathname is 1023 bytes.
Managing Security Auditing of OSS Shell Commands For example, assume that a fileset is mounted on /usr/src/projects/mine. The following sequence of calls occur: rename("/usr/src/projects", "/usr/src/tasks"); open("/usr/src/tasks/mine/main.c"); The audit record for the open call might contain /usr/src/projects/mine/main.c (the old pathname) rather than /usr/src/tasks/mine/main.c (the new pathname).
Managing Security • Use of suid Scripts The password for the user ID of a new shell Security is enforced by requiring the user to complete a normal login procedure for the new login name. The new user ID stays in force until the shell exits. The new password stays in force until it is changed again. For super ID users, the shell substitutes a # (number sign) for its usual prompt. You can specify a login shell using /bin/sh with the appropriate environment variables.
Managing Security Preventing Security Problems suid script files have permissions that either start with the digit 4 or have “s” as the owner’s execution permission bit, as shown in the following example: $ ls -l dfile -rwxr-xr-x 1 MANUALS.HENRYP $ chmod 4755 dfile $ ls -l dfile -rwsr-xr-x 1 MANUALS.HENRYP MANUALS 29 Jun 27 15:24 dfile MANUALS 29 Jun 27 15:24 dfile You can detect suid script files by using the find command with the -perm flag, as shown in the following example: $ find . .
Managing Security • • Identifying Attempts to Break Security Do not provide network services servers such as rshd where mechanisms such as the hosts.equiv file can be used to bypass Safeguard validations. For remote procedure call (RPC) applications, encourage your developers to use the application program interface provided by the HP NonStop Distributed Computing Environment (DCE).
Managing Security Identifying Attempts to Break Security Open System Services Management and Operations Guide— 527191-002 8- 30
9 Managing With the Shell The shell is the interactive interface to the Open System Services (OSS) environment. The OSS shell is a UNIX Korn shell. This section describes how to set up the shell to best serve your users. Information about using the shell is in the Open System Services User’s Guide. Reference information is in the sh(1)reference page either online or in the Open System Services Shell and Utilities Reference Manual.
Managing With the Shell Customizing the OSS Shell Customizing the OSS Shell You can customize the OSS shell by providing a custom default .profile file for your users, setting up the /etc/profile file to meet your needs, and using the localization features in both of these files. Setting Up a Default .profile File Each user’s environment can be set up by means of a .profile file. This file is executed automatically every time a user logs in. Although a .
Managing With the Shell Setting Up an /etc/profile File OSS configuration was performed using the OSSSETUP utility, this copy was made automatically.) Figure 9-1. Sample /etc/profile File /etc/profile.
Managing With the Shell Setting Up an /etc/profile File For example, the time values used by an OSS shell default to those of the system that the shell runs on. If you maintain an OSS environment for users in a time zone other than that used for your Guardian environment, you can add the TZ environment variable to /etc/profile to make the time zone for your OSS users appropriate to their location.
Managing With the Shell Localizing Software Localizing Software You can use the localization environment variables in: • • An /etc/profile file to customize the behavior of all compatible applications launched from an OSS shell for a specific locale A default .profile file to customize the behavior of all compatible applications launched from a specific user’s OSS shell for a specific locale You can use a different locale value for each environment variable in either file.
Managing With the Shell Localizing Software genxlt iconv locale mkcatdefs runcat Before using these tools, you must change the following entry in the EDIT file $SYSTEM.SYSTEM.UNISTDH and the OSS text (header) file /usr/include/unistd.h: #undef _POSIX2_LOCALEDEF /* no support for the creation of locales */ to say: #define _POSIX2_LOCALEDEF 1 /* New locales allowed */ This change might need to be made again after any software product revision or release version update is installed.
Managing With the Shell Localizing Reference Pages Table 9-2. Locale Names and Filenames (page 2 of 2) Language of locale Filename for locale definition Portuguese pt_PT.ISO8859-1 Spanish es_ES.ISO8859-1 Swedish sv_SE.ISO8859-1 Taiwanese, EUC zh_TW.eucTW Turkish tr_TR.ISO8859-9 Localizing Reference Pages HP provides a set of reference pages appropriate for the en_US.ISO8859-1 locale in the default MANPATH environment variable value /usr/share/man.
Managing With the Shell Slow Performance Slow Performance Slow performance might be the result of many processes left running that are no longer being used. One way to check is to enter an OSS command such as: ps | sort -nr +2 This command lists the output of the ps command in reverse order by the TIME field. The processes that have run for the longest time are at the top of the list.
Managing With the Shell Defragmenting Disks This command closes stdin, directs stdout to the file specified as errlog, directs stderr to the file specified as outlog, then runs the specified command. Here are some examples of using the find command to locate and delete files: • Delete all files in /var/tmp whose names begin with TMP that have not been accessed in thirty days: find /var/tmp -name 'TMP*' -atime -30 | xargs rm • Delete all files that are named a.out or whose names end in .
Managing With the Shell Executing Remote Shell Commands If you have the appropriate privileges, pack preserves the access modes, access and modification dates, and owner of the original file. (For details about these attributes, see the chmod(1) reference page either online or in the Open System Services Shell and Utilities Reference Manual.) Otherwise, pack compresses the file and assigns your owner and group ID to the new file.
Managing With the Shell Parsing Command Options With the getopts Command Parsing Command Options With the getopts Command The getopts command is used only in shell scripts, not at the shell prompt. It parses command options and checks a specified command for legal options. Each time it is invoked, getopts places the next option letter it finds into a variable name that you specify. You can also specify letters that the getopts command is to recognize as valid option values and an option argument to parse.
Managing With the Shell Parsing Command Options With the getopts Command Open System Services Management and Operations Guide— 527191-002 9- 12
10 Managing OSS Devices This section briefly discusses Open System Services (OSS) devices and describes how to manage printers in the OSS environment. The Scope of OSS Device Management The only devices you can manage in the OSS environment are printers. All other devices are managed through the Guardian environment. In the OSS environment, printer and tape-drive definitions are not stored in the /dev directory. Do not modify the contents of the /dev directory.
Managing OSS Devices Specifying a Default Printer Printer management in the OSS environment consists of defining aliases for Guardian spooler-location names. The aliases in the /etc/printcap file provide a systemwide set of definitions. At least one alias should be provided in that file. A user can also define aliases in his or her own printcap file. When you add a new user to the system, consider creating a printcap file in the user’s initial working directory.
Managing OSS Devices Using the /etc/printcap or printcap File Using the /etc/printcap or printcap File Both the /etc/printcap and printcap files contain a list of aliases and the corresponding Guardian spooler-location names for printers. A valid /etc/printcap or printcap file must contain at least one entry to establish the default printer. In the sample file /etc/printcap.
Managing OSS Devices Using the /etc/printcap or printcap File and printcap contains default \FORTY.$S1.#BOOK If the user enters the following OSS shell command: lp file the file file is printed at the Guardian spooler-location named \COMM.$S.#DEFAULT, because that command does not specify a printer definition and therefore uses the default definition in /etc/printcap.
11 Managing Problems Most operational problems are easily resolved by following the recovery recommendations listed in Appendix A, Messages. However, some of the messages indicate that a problem should be reported to HP. This section discusses that possibility. Problem-Reporting Procedures Your site should have a formal procedure for reporting problems detected in its own software or in HP software.
Managing Problems Gathering Version Information About OSS Files When a problem occurs with an executable file, the method you use to obtain productversion information depends on the kind of file involved.The possible methods are: • • For a product with a Subsystem Control Facility (SCF) module, such as the OSS Monitor, you can use the SCF VERSION SUBSYS, VERSION MON, and VERSION PROCESS Commands on page 12-86. For other files, you can use the Guardian VPROC utility.
Managing Problems Gathering Version Information About OSS Files This command produces a display similar to the following: VPROC-T9617D31-(14 APR 95) SYSTEM \NODE Date 18 DEC 1995, 11:21:14 COPYRIGHT TANDEM COMPUTERS INCORPORATED 1986 - 1995 /bin/sh Binder timestamp: Version procedure: Target CPU: AXCEL timestamp: 14MAR95 16:32:52 T8626D30_26MAY95_OSSUTL_AAB UNSPECIFIED 14MAR95 16:33:22 The same information appears if you use the OSS version of the vproc command, as shown in the following example: vproc
Managing Problems Gathering Version Information About OSS Files Open System Services Management and Operations Guide— 527191-002 11- 4
12 Open System Services Monitor The Open System Services (OSS) Monitor enables you to perform operations on filesets, OSS servers, and itself. Those operations are described in detail in Section 5, Managing Filesets, Section 4, Managing Servers, and Section 2, Operating the OSS Environment.
Open System Services Monitor OSS Monitor Features Online Help Facility The OSS PM includes a context-sensitive online help facility similar to that provided in other SCF product modules.
Open System Services Monitor OSS Monitor Features Figure 12-2.
Open System Services Monitor OSS Monitor Features Figure 12-4. SCF Help OSS START FILESET Display The START FILESET command makes an existing fileset available to OSS users. This action is also known as mounting a fileset. Syntax of START FILESET: START [ /OUT / ] FILESET . . . Considerations for START FILESET: . . .
Open System Services Monitor OSS Monitor Features Fault Tolerance and Continuous Availability All servers controlled through the OSS Monitor (except for the $ZTAnn transport agent server) can be run as fault-tolerant process pairs. The OSS Monitor itself can be run as a persistent process, as described in Starting the OSS Monitor as a Persistent Process on page 2-9.
Open System Services Monitor OSS Monitor SCF Command Reference Information If the OSS Monitor encounters a serious internal inconsistency, it attempts to issue the “internal error” message (message E00001). For additional information about OSS messages, see Appendix A, Messages. Subsystem and Process Attributes For the OSS Monitor: • • • • The subsystem ID is 143. The device type is 24. The device subtype is 0. The name of the program file is $SYSTEM.SYSnn.OSSMON.
Open System Services Monitor OSS Monitor SCF Command Reference Information VERSION SUBSYS, VERSION MON, and VERSION PROCESS Commands on page 12-86 Open System Services Management and Operations Guide— 527191-002 12- 7
Open System Services Monitor ADD FILESET Command ADD FILESET Command The ADD FILESET command adds a fileset to the configuration administered through the OSS Monitor. This command affects the contents of the ZOSSFSET file. The syntax of the ADD FILESET command is: ADD [ /OUT filename1/ ] FILESET [$ZPMON.
Open System Services Monitor ADD FILESET Command CATALOG volume specifies the name of the Guardian disk volume to contain the catalog files for the fileset. Disk volume names are not case-sensitive. The volume specified should be mirrored. If the BUFFERED CREATE option is specified, any value specified for the POOL parameter is ignored and the disk volume specified for the CATALOG parameter is used as the creation pool.
Open System Services Monitor ADD FILESET Command OFF Security-sensitive operations against this fileset are not logged in the Safeguard security audit trail. This is the default value. BUFFERED { CREATE | LOG | NONE } specifies the amount of catalog write buffering used for the fileset. CREATE When a request or transaction requires a write to the PXINODE or PXLINK file, buffer the corresponding write to the PXLOG file. Also use the fast file-creation feature for writing new file labels.
Open System Services Monitor ADD FILESET Command DEVICELABEL devicelabel specifies the device label assigned to this fileset. The specified label must not already be in use by another fileset in the database. The device label has 6 characters (the first of which is always zero) and consists of numeric characters (0 - 9) and uppercase letters (A - Z), excluding E, I, O, and U. The root fileset always has a device label of 000000. Valid device labels are in the range 000001 through 0ZZZZZ.
Open System Services Monitor ADD FILESET Command NFSPOOL kbytes specifies the number of kilobytes that the OSS name server uses for buffers for nonretryable Network File System (NFS) operations for the fileset. Possible values are 4 through 128 kilobytes. The default value is 16 kilobytes. See the Open System Services NFS Management and Operations Guide for more information about NFS.
Open System Services Monitor ADD FILESET Command INFO FILESET command might not appear to be the same as the specified pathname. • The MAXDIRTYINODETIME option is meaningful only for filesets that use the BUFFERED LOG option. A fileset with an updated (flushed) inode cache is considered “clean” instead of “dirty” and does not need recovery after a failure. The more often the inode cache is flushed, the less likely a fileset is to be corrupted by a failure and to need recovery after the failure.
Open System Services Monitor • ADD FILESET Command To add the fileset USER2 with the storage-pool file ZOSSPOOL using the volume catalog on $DATA2 and the mount point /user2, enter the following command: ADD /OUT $S.#USR/ FILESET $ZPMON.USER2, CATALOG $DATA2, & POOL ZOSSPOOL, MNTPOINT "/user2", NAMESERVER #ZPNS, & NFSTIMEOUT 60, BUFFERED NONE This command: • ° ° ° Sends informational messages to the spooler location $S.#USR.
Open System Services Monitor • ADD SERVER Command ° Specifies that the inode cache for the fileset USER3 should be flushed to disk after approximately 10 seconds to make its recovery faster by decreasing the time when the cache is considered dirty. ° Performs automatic recovery of the fileset USER3 using a copy of the FSCK program in processor 5 after a shutdown or failure. The recovered fileset is left in a started state. ° Sends FSCK output after an automatic recovery to the spooler $S.#USER3.
Open System Services Monitor ADD SERVER Command servername specifies the name of the server to add. servername must begin with a pound sign (#) followed by a letter, which can be followed by 0 through 5 letters or digits. Only names not currently defined in the ZOSSSERV file are valid values. Server names are not case-sensitive. The $ZPMON prefix can be omitted if you have previously specified $ZPMON in an SCF ASSUME command. TYPE NAME specifies that an OSS name server is being added.
Open System Services Monitor ADD SERVER Command The current persistence count for each server is incremented when that server terminates abnormally or is stopped by something other than SCF. The current persistence count is not incremented because of processor failure. When the AUTORESTART option is omitted from the ADD SERVER command, the default specification is AUTORESTART 3. BACKUPCPU backup_processor specifies the processor number of the processor where the backup server process will run.
Open System Services Monitor ADD SERVER Command If the BACKUPCPUOK attribute for the server is TRUE and the primary processor does not become available after the specified period has elapsed, the OSS Monitor attempts to start the server in its backup processor. When the MAXWAITTIME option is omitted from the ADD SERVER command, the default specification is MAXWAITTIME 0. SQLTIMEOUT seconds specifies the number of seconds an OSS name server waits for a response from SQLCAT to a request.
Open System Services Monitor ALTER FILESET Command ALTER FILESET Command The ALTER FILESET command changes the configuration of a fileset administered through the OSS Monitor. The information entered in the command is added to or altered in the ZOSSFSET file. The syntax of the ALTER FILESET command is: ALTER [ /OUT filename1/ ] FILESET [$ZPMON.
Open System Services Monitor ALTER FILESET Command You can specify the following options in any order. You must specify at least one option: AUDITENABLED { ON | OFF } specifies whether the fileset is audited (ON) or not audited (OFF). ON Safeguard global auditing controls govern which objects are audited. OFF Security-sensitive operations against this fileset are not logged in the Safeguard security audit trail. Only the super ID (255,255) can change the value of AUDITENABLED.
Open System Services Monitor ALTER FILESET Command CATALOG volume specifies the name of the Guardian disk volume to contain the catalog files for the fileset. Any existing catalog files are moved to this disk volume; the subvolume of the catalog files is unchanged. Disk volume names are not case-sensitive.
Open System Services Monitor ALTER FILESET Command MAXDIRTYINODETIME seconds2 specifies the approximate number of seconds that cached inodes of the fileset remain in the OSS name server’s inode cache without being updated. seconds2 must be a value in the range 1 through 600. When the MAXDIRTYINODETIME option is omitted, the previous value for the fileset is unchanged. MAXINODES maxinodes specifies the approximate maximum number of inodes that can be created for the fileset.
Open System Services Monitor ALTER FILESET Command NFSTIMEOUT seconds specifies the number of seconds that the OSS name server retains the results of nonretryable Network File System (NFS) operations for the fileset. Valid values are in the range 60 through 300. If the NFSTIMEOUT option is omitted, the timeout interval for the fileset is not changed. POOL filename2 specifies the Guardian filename of the storage-pool file that contains the volume list to use for file creation for the fileset.
Open System Services Monitor ALTER FILESET Command When the REPORT option is omitted, the previous value for the fileset is unchanged. Considerations • • • • • • The ALTER FILESET command can be used only by super-group users (255,nnn). You can use the ALTER FILESET command on a fileset that is not in the STOPPED state. However, the changes do not take effect until the fileset is stopped and restarted. Assigning a new storage-pool file to a fileset has no effect on existing OSS files in the fileset.
Open System Services Monitor ALTER FILESET Command Examples • To alter the read-write fileset USER1 to be a read-only fileset using the volume catalog on $DATA1 and the mount point /user1, enter the following command: ALTER /OUT CMDLOG/ FILESET $ZPMON.USER1, BUFFERED LOG, & CATALOG $DATA1, & MNTPOINT "/user1", NAMESERVER #ZPNS1, NFSTIMEOUT 60, & READONLY TRUE This command: • ° ° Sends informational messages to the file CMDLOG. ° ° ° Specifies that $DATA contains the catalog files for the fileset.
Open System Services Monitor ALTER SERVER Command ALTER SERVER Command The ALTER SERVER command changes the configuration of a server in the set of servers configured for administration through the OSS Monitor. The information entered in the command is added to or altered in the ZOSSSERV file. This command is not valid for OSS transport agent servers.
Open System Services Monitor ALTER SERVER Command The $ZPMON prefix can be omitted if you have previously specified $ZPMON in an SCF ASSUME command. You can specify the following options in any order. You must specify at least one option: AUTORESTART ntimes specifies the persistence count of the server, where ntimes is an integer value in the range 0 through 10. The persistence count is the number of times the OSS Monitor will automatically restart the server during a 10-minute period.
Open System Services Monitor ALTER SERVER Command If the BACKUPCPU option is omitted, the processor for the backup server process is unchanged. BACKUPCPUOK { TRUE | FALSE } specifies whether the server should be started in its configured backup processor when its configured primary processor is unavailable and the server is being restarted automatically.
Open System Services Monitor ALTER SERVER Command LINKCACHE size2 specifies the number of entries permitted in the link cache used by the OSS name server. Possible values are in the range 128 through 500000. The functional upper limit might be smaller than the theoretical maximum, because the upper limit depends upon the amount of memory available when the OSS name server is started. If the LINKCACHE option is omitted, the number of entries permitted in the link cache is unchanged.
Open System Services Monitor ALTER SERVER Command If the MSGMQB option is omitted, the maximum number of bytes allowed is unchanged. The default value is the maximum, 65535. This option is valid only for OSS message-queue servers. MSGMSIZE msgmsize specifies the maximum number of bytes allowed for a message. Valid values are in the range 1 through 32000. If the MSGMSIZE option is omitted, the maximum number of bytes allowed is unchanged. The default value is the maximum, 32000.
Open System Services Monitor • • ALTER SERVER Command The INODECACHE, LINKCACHE, and SQLTIMEOUT parameters are not valid when the OSS message-queue server is specified in an ALTER SERVER command. OSS transport agent servers cannot be modified and are not a valid server type for the ALTER SERVER command.
Open System Services Monitor ALTER SUBSYS, ALTER MON, and ALTER PROCESS Commands ALTER SUBSYS, ALTER MON, and ALTER PROCESS Commands The ALTER SUBSYS, ALTER MON, and ALTER PROCESS commands all change the configuration of the OSS Monitor. The information entered in the command is added to or altered in the ZOSSPARM file.
Open System Services Monitor ALTER SUBSYS, ALTER MON, and ALTER PROCESS Commands You can specify the following options in any order. You must specify at least one option: AUTOSTART { AUTO | MANUAL } specifies whether the automatic startup service is enabled. AUTO All filesets and servers that are configured to use the automatic startup service are automatically restarted after a system load or restart. MANUAL Filesets and servers are not automatically restarted after a system load or restart.
Open System Services Monitor ALTER SUBSYS, ALTER MON, and ALTER PROCESS Commands configured report file for the fileset is unavailable. The specified value must be either NULL, a blank, or the Guardian filename of a spooler location. If filename2 is not available when it is needed, reports are created as if NULL had been specified. The initial value (set when the configuration database is created) is blank.
Open System Services Monitor • CONTROL FILESET Command The AUTOSTART attribute of the subsystem can also be set by starting the OSS Monitor with a PARAM value or RUN command line parameter of AUTOSTART AUTO or AUTOSTART MANUAL. The value entered in the PARAM or command line takes precedence over the previously configured value for the subsystem. The precedence is, in descending order: 1. Command line parameter 2. PARAM specification 3.
Open System Services Monitor CONTROL FILESET Command FILESET [$ZPMON.]filesetname specifies the name of the OSS fileset to be altered. Only the names of currently started filesets defined in the ZOSSFSET file are valid values. Fileset names are not case-sensitive. The $ZPMON prefix can be omitted if you have previously specified $ZPMON in an SCF ASSUME command. You can specify the following options in any order.
Open System Services Monitor CONTROL SERVER Command Considerations • • • • The CONTROL FILESET command can be used only by super-group users (255,nnn). When you use the SYNC option and the fileset has the READONLY TRUE attribute, the BUFFERED attribute value in the ZOSSFSET file is ignored and the OSS name server uses the BUFFERED NONE attribute.
Open System Services Monitor CONTROL SERVER Command The syntax of the CONTROL SERVER command is: CONTROL [ /OUT filename1/ ] SERVER server_processname { [ , RESET { ALL | MAXINODECACHEUSED | MAXLINKCACHEUSED } ] [ , SYNC ] } OUT filename1 specifies the name of a Guardian output file for informational messages. You can either read this file with a text editor or display it with the FUP COPY command. SERVER server_processname specifies the server being altered.
Open System Services Monitor CONTROL SERVER Command The following OSS name server attributes are updated as soon as their ZOSSSERV values are changed with the SCF ALTER SERVER command and do not require use of the CONTROL SERVER command SYNC option: AUTORESTART BACKUPCPUOK MAXWAITTIME If you omit the SYNC option, the values in use for the server are not updated from the ZOSSSERV file. Considerations • • • The CONTROL SERVER command can be used only by super-group users (255,nnn).
Open System Services Monitor DELETE FILESET Command DELETE FILESET Command The DELETE FILESET command removes an OSS fileset from the configuration administered through the OSS Monitor. This command affects the contents of the ZOSSFSET file. The syntax of the DELETE FILESET command is: DELETE [ /OUT filename/ ] FILESET [$ZPMON.]filesetname OUT filename specifies the name of a Guardian output file for informational messages.
Open System Services Monitor DELETE SERVER Command DELETE SERVER Command The DELETE SERVER command removes an OSS name server, except the root OSS name server, from the configuration administered through the OSS Monitor. This command does not apply to any of the other OSS servers. The syntax of the DELETE SERVER command is: DELETE [ /OUT filename / ] SERVER server_processname OUT filename specifies the name of a Guardian output file for informational messages.
Open System Services Monitor DIAGNOSE FILESET Command DIAGNOSE FILESET Command The DIAGNOSE FILESET command causes the Guardian FSCK utility to check the integrity of an existing fileset. The fileset must be stopped (unmounted) before it can be checked. The syntax of the DIAGNOSE FILESET command is: DIAGNOSE [ /OUT filename1 / ] FILESET [$ZPMON.
Open System Services Monitor DIAGNOSE FILESET Command OPTION { START | STOP } starts or stops a diagnosis by the FSCK utility. START Starts a new diagnosis and sets the state of the fileset to DIAGNOSING. STOP Stops a diagnosis that is in progress and sets the state of the fileset to STOPPED. If this option is omitted, the default value is OPTION START.
Open System Services Monitor DIAGNOSE FILESET Command NONE Do not repair any inconsistencies. This is the default specification. OPEN Same as SERIOUS, but also assumes that there might be open files in this fileset. This option indicates that inodes cannot be reused until the next unmount or remount. SERIOUS Repair only serious inconsistencies. This option can speed up the execution of a fileset check. If the REPAIR option is omitted, the default value is REPAIR NONE.
Open System Services Monitor • • • • DIAGNOSE FILESET Command You can use the SCF STATUS FILESET command to determine whether the integrity check is complete. If FSCK fails, the fileset is placed in the UNKNOWN state. The FSCK utility checks for the inconsistencies listed in Table 5-2 on page 5-28. The DIAGNOSE FILESET command can be used only by super-group users (255,nnn).
Open System Services Monitor INFO FILESET Command INFO FILESET Command The INFO FILESET command displays information about the attributes of an OSS fileset. The information displayed is from the ZOSSFSET and ZOSSSERV files. The syntax of the INFO FILESET command is: INFO [ /OUT filename/ ] FILESET [$ZPMON.]filesetname [ , DETAIL ] [ , SEL [ NOT ] state ] OUT filename specifies the name of a Guardian output file for informational messages.
Open System Services Monitor INFO FILESET Command The information returned for a summary report has the following format (an asterisk indicates an attribute whose value can be changed using an SCF command): OSS Info FILESET fileset-devicename FilesetName filesetname *MntPoint directory fileset-devicename is the name of the fileset whose information is displayed, shown as a device of the OSS Monitor process. filesetname is the name of the fileset whose information is being displayed.
Open System Services Monitor INFO FILESET Command volume is the name of the Guardian disk volume that contains the catalog files for the fileset. servername1 is the server name of the OSS name server that administers the fileset. choice indicates the amount of catalog write buffering used for the fileset. Value Meaning CREATE Buffer catalog writes during file creation within the fileset. This is the fast-create option. LOG Buffer only log file activity. NONE Do not buffer anything.
Open System Services Monitor INFO FILESET Command user-name2 is the user name of the user who last modified the configuration or state of the fileset. time2 is the timestamp for the last time that the fileset configuration was changed, in the form dd mmm yyyy hh:mm:ss.mil. filesetname is the name of the fileset on which this fileset is mounted. servername2 is the server name of the OSS name server for the fileset identified by filesetname. directory is the OSS pathname of the mount point for the fileset.
Open System Services Monitor INFO FILESET Command maxinodes is the approximate maximum number of inodes allowed for the fileset. If a fileset has not yet been started after installing the first RVU that supports this feature, the word unknown is displayed. Example To obtain a detailed report of the configuration of the fileset USER1 and send informational messages to the file CMDLOG, enter the following command: INFO /OUT CMDLOG/ FILESET $ZPMON.
Open System Services Monitor INFO SERVER Command INFO SERVER Command The INFO SERVER command displays information about a server administered by the OSS Monitor. The information displayed is from the ZOSSSERV file. The syntax of the INFO SERVER command is: INFO [ /OUT filename/ ] SERVER server_processname [ , DETAIL ] [ , SEL [ NOT ] state ] OUT filename specifies the name of a Guardian output file for informational messages.
Open System Services Monitor INFO SERVER Command SEL [ NOT ] state directs the command to apply only to servers that are in the specified state or, when the NOT field is used, only to servers that are not in the specified state. Valid values for state are: Value Meaning STARTED STOPPED The server is started. The server is ready to be started.
Open System Services Monitor INFO SERVER Command recreated its backup process in the next available processor, as implemented in the OSS message-queue server. The information returned for a detailed report has the following format (an asterisk indicates an attribute whose value can be changed using an SCF command): OSS Detailed Info SERVER server-devicename Type.................. *CPU................... *BackupCPU............. CreateBy.............. CreateTime............ AlterBy............... AlterTime.
Open System Services Monitor INFO SERVER Command user-name1 is the user name of the user or process that created the initial configuration of the server. time1 is the timestamp for the time when the initial configuration of the server was created, in the form dd mmm yyyy hh:mm:ss.mil. user-name2 is the user name of the user who last modified the configuration or state of the server. time2 is the timestamp for the last time that the server configuration was changed, in the form dd mmm yyyy hh:mm:ss.mil.
Open System Services Monitor INFO SERVER Command maxmsg specifies the maximum number of messages allowed on all OSS message queues on a system. Valid values are in the range 1 through 16384. This parameter is valid only for OSS message-queue servers. msgmsize specifies the maximum number of bytes allowed for a message. Valid values are in the range 1 through 32000. This parameter is valid only for OSS message queue-servers.
Open System Services Monitor INFO SERVER Command A display such as the following is written to CMDLOG: OSS Detailed Info SERVER \NODE1.$ZPMON.#ZPNS Type.................... *CPU..................... *BackupCPU............... CreateBy................ CreateTime.............. AlterBy................. AlterTime............... *Params.................. *BackupCPUOK............. *MaxWaitTime............. *DesiredState............ *AutoRestart............. PersistenceCount........
Open System Services Monitor INFO SUBSYS, INFO MON, and INFO PROCESS Commands INFO SUBSYS, INFO MON, and INFO PROCESS Commands The INFO SUBSYS, INFO MON, and INFO PROCESS commands all display information about the OSS subsystem. The information displayed is from the ZOSSPARM file. The syntax of the INFO SUBSYS, INFO MON, and INFO PROCESS commands is: INFO [ /OUT filename/ ] object-type [ process ] [ , DETAIL ] OUT filename specifies the name of a Guardian output file for informational messages.
Open System Services Monitor INFO SUBSYS, INFO MON, and INFO PROCESS Commands process-filename is the process name of the subsystem whose information is displayed, shown as the Guardian filename of the OSS Monitor process. process_name is the process name of the subsystem whose information is displayed, $ZPMON. seconds is the number of seconds that the OSS Monitor waits for a response from an OSS name server to an OSS Monitor request.
Open System Services Monitor INFO SUBSYS, INFO MON, and INFO PROCESS Commands processor is the processor number of the processor in which the FSCK program runs when the SCF DIAGNOSE FILESET or DELETE FILESET command is used and either no processor is configured for the fileset or the processor configured for the fileset is unavailable. filespec is the Guardian filename or spooler location of the default FSCK log file.
Open System Services Monitor INFO SUBSYS, INFO MON, and INFO PROCESS Commands Example To obtain a detailed report of the configuration of the OSS subsystem and send informational messages to the file CMDLOG, enter the following command: INFO /OUT CMDLOG/ SUBSYS $ZPMON, DETAIL A display such as the following is written to CMDLOG: OSS Detailed Info SUBSYS \NODE1.$ZPMON *IOTimeout............. *FsckCPU............... *Report ............... *ZOSSVol............... CreateBy.............. CreateTime...........
Open System Services Monitor NAMES Command NAMES Command The NAMES command lists objects managed through the OSS Monitor. The syntax of the NAMES command is: NAMES [ /OUT filename/ ] [ object-type ] [ object-name ] [, SEL [ NOT ] state ] OUT filename specifies the name of a Guardian output file to receive the list of objects. You can either read this file with a text editor or display it with the FUP COPY command. object-type specifies the types of objects whose names are to be listed.
Open System Services Monitor NAMES Command For the FILESET object type, valid values for state are: Value Meaning DIAGNOSING STARTED STOPPED UNKNOWN The fileset is being diagnosed by the FSCK utility. The fileset is started (mounted). The fileset is ready to be started (mounted) or diagnosed. The fileset is in an unknown state. The OSS Monitor does not recognize the SEL parameter for object types other than SERVER or FILESET. The information returned has the following format: $ZPMON Names SUBSYS node.
Open System Services Monitor NAMES Command Example To list the names of all OSS filesets, enter the following SCF commands: ASSUME $ZPMON NAMES FILESET * The following information is displayed: OSS Names FILESET \NODE1.$ZPMON.
Open System Services Monitor RENAME FILESET Command RENAME FILESET Command The RENAME FILESET command changes an existing fileset name. The syntax of the RENAME FILESET command is: RENAME [ /OUT filename/ ] FILESET [$ZPMON.]filesetname, [$ZPMON.]newname filename specifies the name of a Guardian output file for informational messages. You can either read this file with a text editor or display it with the FUP COPY command. FILESET [$ZPMON.]filesetname is the name of a fileset to be renamed.
Open System Services Monitor START FILESET Command START FILESET Command The START FILESET command makes an existing OSS fileset available to users (also known as mounting the fileset). The syntax of the START FILESET command is: START [ /OUT filename/ ] FILESET [$ZPMON.]filesetname OUT filename specifies the name of a Guardian output file for informational messages. You can either read this file with a text editor or display it with the FUP COPY command. FILESET [$ZPMON.
Open System Services Monitor • START FILESET Command To start (mount) the ROOT fileset, enter the following SCF commands: ASSUME $ZPMON START FILESET ROOT To start (mount) the fileset USER1 and send informational messages to the file CMDLOG, enter the following command: START /OUT CMDLOG/ FILESET $ZPMON.
Open System Services Monitor START SERVER Command START SERVER Command The START SERVER command starts an OSS server. The syntax of the START SERVER command is: START [ /OUT filename/ ] SERVER server_processname OUT filename specifies the name of a Guardian output file for informational messages. You can either read this file with a text editor or display it with the FUP COPY command. SERVER server_processname specifies the server being started. server_processname has the following form: [$ZPMON.
Open System Services Monitor STATUS FILESET Command STATUS FILESET Command The STATUS FILESET command displays status information about a fileset administered by the OSS Monitor.The syntax of the STATUS FILESET command is: STATUS [ /OUT filename/ ] FILESET [$ZPMON.]filesetname [ , DETAIL ] [ , SEL [ NOT ] state ] OUT filename specifies the name of a Guardian output file for informational messages. You can either read this file with a text editor or display it with the FUP COPY command. FILESET [$ZPMON.
Open System Services Monitor STATUS FILESET Command The information returned for a summary report has the following format: OSS Status FILESET fileset-devicename FilesetName filesetname State state LastError error ErrorDetail error-detail fileset-devicename is the name of the fileset whose status is displayed, shown as a device of the OSS Monitor process. filesetname is the name of the fileset whose status is displayed. state describes the state of the fileset.
Open System Services Monitor STATUS FILESET Command The information returned for a detailed report has this format: OSS Detailed Status FILESET fileset-devicename State.................. state MountTime.............. time1 LastError.............. error ErrorDetail............ error-detail ErrorTime.............. time2 FsckName............... volume LastControlSyncTime.... time3 AlterAfterSyncOrMount.. status1 *AuditEnabled........... status2 *Buffered............... type *ReadOnly...............
Open System Services Monitor STATUS FILESET Command time2 is the timestamp for the time when the last error for the fileset was reported, in the form dd mmm yyyy hh:mm:ss.mil. If no error has been reported, this value is the same as the time1 value. status1 indicates whether the configuration of the fileset has changed after its most recent start or synchronization with ZOSSFSET: Value Meaning FALSE The fileset configuration has not changed. TRUE The fileset configuration has changed.
Open System Services Monitor STATUS FILESET Command type indicates the type of buffering performed for the fileset: Value Meaning CREATE When a request or transaction requires a write to the PXINODE or PXLINK file, the corresponding write is buffered to the PXLOG file. Also uses the fast file-creation feature for writing new file labels. When a request or transaction requires a write to the PXINODE or PXLINK file, the corresponding write is buffered to the PXLOG file.
Open System Services Monitor STATUS FILESET Command number2 is the approximate maximum number of inodes that the OSS name server allows for the fileset. This field is displayed only if the fileset is in the STARTED state. number3 is the current number of inodes that the OSS name server is using for the fileset, displayed as a decimal number and as a percentage of the maximum number allowed. This field is displayed only if the fileset is in the STARTED state.
Open System Services Monitor STATUS FILESET Command volume_list3 lists the names of the disk volumes in the fileset’s storage pool that do not have space available for new file creation. volume_list4 lists the names of the disk volumes in the fileset’s storage pool that currently contain no files and can be safely removed from the storage pool. volume_list5 lists the names of the disk volumes in the storage-pool for the fileset that already contain OSS files.
Open System Services Monitor STATUS FILESET Command Examples • To determine the status of the fileset USER1 and send informational messages to the file CMDLOG, enter the following command: STATUS /OUT CMDLOG/ FILESET $ZPMON.USER1 A display such as the following is written to CMDLOG: OSS Status FILESET USER1 FilesetName USER1 • State STARTED LastError 0 ErrorDetail 0 To see a listing of all information available for the fileset USER1, enter the following command: STATUS /OUT CMDLOG/ FILESET $ZPMON.
Open System Services Monitor STATUS FILESET Command To determine the status of the fileset USER1 and send informational messages to the file CMDLOG, enter the following command: STATUS /OUT CMDLOG/ FILESET $ZPMON.
Open System Services Monitor STATUS SERVER Command STATUS SERVER Command The STATUS SERVER command displays status information about a server administered by the OSS Monitor. The syntax of the STATUS SERVER command is: STATUS [ /OUT filename/ ] SERVER server_processname [ , DETAIL ] [ , SEL [ NOT ] state ] OUT filename specifies the name of a Guardian output file for informational messages. You can either read this file with a text editor or display it with the FUP COPY command.
Open System Services Monitor STATUS SERVER Command SEL [ NOT ] state directs the command to apply only to servers that are in the specified state or, when the NOT field is used, only to servers that are not in the specified state. Valid values for state are: Value Meaning STARTED STOPPED The server is started. The server is ready to be started.
Open System Services Monitor STATUS SERVER Command The information returned for a detailed report has this format: OSS Detailed Status SERVER server-devicename State.................. StartTime.............. LastError.............. ErrorDetail............ ErrorTime.............. LastControlSyncTime.... AlterAfterSyncOrStart.. *CPU.................... *BackupCPU.............. *SQLTimeout............. *InodeCache............. InodeCacheInUse........ MaxInodeCacheUsed...... TimeMaxInodeCacheUsed..
Open System Services Monitor STATUS SERVER Command time2 is the timestamp for the time when the last error for the server was reported, in the form dd mmm yyyy hh:mm:ss.mil. If no error has been reported, this value is the same as the time1 value. The following fields are displayed only when an OSS name server is in the STARTED state: time3 is the timestamp for the time when the OSS name server was last synchronized with ZPOSFSERV, in the form dd mmm yyyy hh:mm:ss.mil.
Open System Services Monitor STATUS SERVER Command time5 is the timestamp for the time when the OSS name server last reached its highwater mark for inode cache use, in the form dd mmm yyyy hh:mm:ss.mil. time6 is the timestamp for the time when the OSS name server last had its highwater mark for inode cache use reset, in the form dd mmm yyyy hh:mm:ss.mil. size4 indicates the number of links that the OSS name server is allowed to cache.
Open System Services Monitor STATUS SERVER Command A display similar to the following is returned: OSS Detailed Status SERVER \NODE1.$ZPMON.#ZPNS State....................... STARTED StartTime................... 25 Nov 2002, 10:23:20.123 LastError................... 0 ErrorDetail................. 0 ErrorTime................... 25 Nov 2002, 10:23:20.123 LastControlSyncTime........ AlterAfterSyncOrStart...... *CPU........................ *BackupCPU.................. *SQLTimeout.................
Open System Services Monitor STOP FILESET Command STOP FILESET Command The STOP FILESET command makes an existing, started OSS fileset unavailable to users (also known as unmounting the fileset). The syntax of the STOP FILESET command is: STOP [ /OUT filename/ ] FILESET [$ZPMON.]filesetname OUT filename specifies the name of a Guardian output file for informational messages. You can either read this file with a text editor or display it with the FUP COPY command. FILESET [$ZPMON.
Open System Services Monitor STOP SERVER Command STOP SERVER Command The STOP SERVER command stops an OSS server. The syntax of the STOP SERVER command is: STOP [ /OUT filename/ ] SERVER server_processname OUT filename specifies the name of a Guardian output file for informational messages. You can either read this file with a text editor or display it with the FUP COPY command. SERVER server_processname specifies the server being stopped. server_processname has the following form: [$ZPMON.
Open System Services Monitor • STOP SERVER Command To stop all OSS sockets local servers, OSS message-queue servers, and OSS transport agent servers and send informational messages to the file CMDLOG, enter the following command: STOP /OUT CMDLOG/ SERVER * Open System Services Management and Operations Guide— 527191-002 12 -85
Open System Services Monitor VERSION SUBSYS, VERSION MON, and VERSION PROCESS Commands VERSION SUBSYS, VERSION MON, and VERSION PROCESS Commands The VERSION SUBSYS, VERSION MON, and VERSION PROCESS commands all display product-version information for a specified OSS object known to SCP.
Open System Services Monitor VERSION SUBSYS, VERSION MON, and VERSION PROCESS Commands subsystem-name is the name of the SCF subsystem associated with the object. version-information-banner provides the product-version information for the object. See Section 11, Managing Problems, for more information about product-version information and the related Guardian VPROC utility.
Open System Services Monitor VERSION SUBSYS, VERSION MON, and VERSION PROCESS Commands Considerations • • The product-version information for the SCF Kernel subsystem is displayed as part of the SCF startup banner and when the SCF ENV command is used. In the SCF object hierarchy, SUBSYS is the highest of the possible objects for this command. Examples • The command VERSION $ZPMON produces output such as the following: VERSION SUBSYS \NODE.
A Messages Status, warning, and error messages are sent to the console by: • • • • • • The Open System Services (OSS) EasySetup utilities. These messages are described under OSS EasySetup Utility Messages on page A-2. The Guardian Catalog Volume Tool (CVT) utility. These messages are described under CVT Messages on page A-3. The FSCK fileset integrity checker when the Subsystem Control Facility (SCF) DELETE FILESET or DIAGNOSE FILESET command is used.
Messages OSS EasySetup Utility Messages numbers (for example, E-00019); only messages specific to the OSS subsystem with positive message numbers (for example, E00019) are described in this guide. For further information about EMS messages, see the Operator Messages Manual. OSS EasySetup Utility Messages The following console messages can appear on the terminal that issued an EasySetup command.
Messages CVT Messages Recovery. If EMS events are needed, install or reinstall the missing SPI file, carefully following the instructions for DSM/SCM use. CVT Messages The following warning and error messages appear on the terminal that issued the CVT command. The Guardian CVT utility does not generate Event Management Service (EMS) events. CVT Warning Message CVT issues the following warning message. Warning:filename - No such File filename is a Guardian filename used in the command. Cause.
Messages CVT Error Messages filename identifies the affected file. Cause. The Guardian FILE_OPEN_ procedure call could not be completed. Effect. CVT did not open the indicated file. Recovery. Check the command and correct it as necessary. FILE_PURGE_ Error error on file filename error is a Guardian file-system error value returned by the underlying Guardian FILE_PURGE_ procedure call. See the Guardian Procedure Errors and Messages Manual for information about the indicated error.
Messages CVT Error Messages Cause. There is a PXCKSTAT file on one of the subvolumes involved in processing the command. CVT does not move catalogs to, from, or within a subvolume that contains a PXCKSTAT file—nor does it purge a saved catalog from such a subvolume. Effect. CVT terminates. Recovery. Run the FSCK utility against the catalog before attempting the CVT command again. ***Incomplete Command*** Cause. The command entered was incomplete. Effect. CVT terminates. Recovery.
Messages FSCK Messages Effect. CVT terminates. Recovery. List the files in the subvolume to see the FSN values used in their Guardian file IDs. Try the command again with an FSN from those shown in the file IDs. See Generated Catalog Files on page 5-31 for more information about FSN values. ***Invalid Subvolume Name*** - token token identifies the invalid subvolume name. Cause. The value token followed the keyword IN in the command line, but it is not a valid Guardian subvolume name. Effect.
Messages FSCK Consistent-Fileset Messages FSCK Consistent-Fileset Messages If the FSCK utility finds no relevant inconsistencies or errors in a fileset, it issues one of the following messages (listed in alphabetic order). After issuing one of these messages, FSCK clears the mounted flag and the bitmap indicating the types of inconsistencies previously detected by the OSS name server. *** CATALOG AND ZYQ SUBVOLUMES HAVE BEEN PURGED *** Cause. FSCK successfully deleted a fileset. Effect. None. Recovery.
Messages FSCK Inconsistency and Error Messages SERIOUS These messages are issued when FSCK detects a serious inconsistency. A catalog integrity problem exists. WARNING These messages are issued when a recoverable error occurs or when FSCK takes an action (such as purging an empty orphan ZYQ file) that is not reported as an inconsistency. WARNING messages report conditions that FSCK encounters or actions that FSCK takes that do not by themselves indicate problems with the catalog.
Messages FSCK Inconsistency and Error Messages Cause. The FSCK utility encountered a file-management error on an operation (given by operation) on the indicated file because of the error described by errornumber and description. Some common examples of this message are: Message Meaning OPEN Error 11 (Record Not Found) on PXINODE There is no OSS fileset catalog in the specified (or implied) catalog subvolume. The ZOSSFSET file specifies an incorrect volume.
Messages FSCK Inconsistency and Error Messages 8 WARNING 8 - FSCK Run Number nnnn was Interrupted nnnn identifies the FSCK serial number (FSN) of the affected invocation of FSCK. Cause. The indicated invocation of FSCK was interrupted before it finished. Effect. A new FSN is not assigned. If FSCK did not finish creating a new catalog, the existing saved catalog files (PXINnnnn, PXLInnnn, and PXLOnnnn) are used.
Messages FSCK Inconsistency and Error Messages 15 WARNING 15 - Catalog Already Downgraded Cause. The FSCK utility was asked to downgrade a catalog to the previous format, but the catalog is already in that format. Effect. FSCK does not downgrade the catalog. Recovery. Informational message only; no corrective action is needed. 16 WARNING 16 - Dirty Catalog using Fast Create; REPAIR ALL will be performed Cause.
Messages FSCK Inconsistency and Error Messages 99 ERROR 99 - HEAP OVERFLOW Cause. The FSCK utility has exhausted its internal memory. Effect. FSCK terminates abnormally. Recovery. Set the SAVEABEND attribute of the FSCK program file using the nld command. Rerun FSCK to produce a saveabend file.
Messages FSCK Inconsistency and Error Messages If the REPAIR SERIOUS, REPAIR OPEN, or REPAIR ALL option was specified in the command, this is an informational message only; no corrective action is needed. 101 SERIOUS 101 - Duplicate Link ID - Parent:parent, Child:child, Name:link linkname parent indicates the inode number of the parent end of the link. child indicates the inode number of the child end of the link. linkname identifies the affected link. Cause.
Messages FSCK Inconsistency and Error Messages Effect. The indicated record is added to the new PXLINK file. The missing link will be for one of the reserved inodes (/, /G, /E, /lost+found, /dev, /dev/tty, /dev/null, or the superblock). Recovery. If the REPAIR option was not specified or the REPAIR NONE option was specified in the Subsystem Control Facility (SCF) DIAGNOSE FILESET command, the catalog is still in an inconsistent state. Reissue the command with the REPAIR SERIOUS or REPAIR OPEN option.
Messages FSCK Inconsistency and Error Messages catalog is still in an inconsistent state. Reissue the command with the REPAIR SERIOUS or REPAIR OPEN option. If the REPAIR SERIOUS, REPAIR OPEN, or REPAIR ALL option was specified in the command, this is an informational message only; no corrective action is needed. 202 SERIOUS 202 - Corrupt Inode, Inode=inode-number inode-number indicates the affected inode. Cause. The FSCK utility has detected a corrupt record in the PXINODE file. Effect.
Messages FSCK Inconsistency and Error Messages 205 MINOR 205 - Missing ZYQ File, Inode=inode-number inode-number indicates the affected inode. Cause. The indicated regular file inode has no corresponding ZYQ file. Effect. The inode and any links to it are omitted from the new catalog. Recovery.
Messages FSCK Inconsistency and Error Messages Recovery. If the REPAIR option was not specified or the REPAIR NONE option was specified in the Subsystem Control Facility (SCF) DIAGNOSE FILESET command, the catalog is still in an inconsistent state. Reissue the command with the REPAIR ALL option. If the REPAIR SERIOUS, REPAIR OPEN, or REPAIR ALL option was specified in the command, this is an informational message only; no corrective action is needed.
Messages FSCK Inconsistency and Error Messages For the SERIOUS message: Cause. A record with the indicated invalid inode number as a key was found in the current PXINODE file. The inode number is less than or equal to zero or is greater than 231. Effect. The record is omitted from the new PXINODE file. Recovery. If the REPAIR option was not specified or the REPAIR NONE option was specified in the Subsystem Control Facility (SCF) DIAGNOSE FILESET command, the catalog is still in an inconsistent state.
Messages FSCK Inconsistency and Error Messages 211 SERIOUS 211 - Loop in Directory Graph, Inode=inode-number inode-number indicates the affected inode. Cause. The directory graph forms a loop that includes the indicated inode. The inode is linked through other inodes back to itself; this means that the directory referred to by the inode number is its own parent. Effect. The inode is unlinked and placed in the OSS /lost+found directory. Recovery.
Messages FSCK Inconsistency and Error Messages Effect. A currently unused inode is allocated, the ZYQ file is renamed to correspond to the new inode, and the inode is placed in the OSS /lost+found directory. Recovery. If the REPAIR ALL option was not specified in the Subsystem Control Facility (SCF) DIAGNOSE FILESET command, the catalog is still in an inconsistent state. Reissue the command with the REPAIR ALL option.
Messages FSCK Inconsistency and Error Messages If the REPAIR ALL option was specified in the command, this is an informational message only; no corrective action is needed. 216 MINOR 216 - ZYQ File Conflict - filename filename indicates the affected file. Cause. The inode corresponding to the indicated file does not refer to that file. Effect. A currently unused inode is allocated, the ZYQ file is renamed to correspond to the new inode, and the inode is placed in the OSS /lost+found directory. Recovery.
Messages FSCK Inconsistency and Error Messages If the REPAIR SERIOUS, REPAIR OPEN, or REPAIR ALL option was specified in the command, this is an informational message only; no corrective action is needed. 219 WARNING 219 - Root Fileset Cause. The FILESET ROOT option was not specified in the Subsystem Control Facility (SCF) DIAGNOSE FILESET command, but the fileset catalog is a root catalog. Effect. The FSCK utility does not check for the existence of those files that must be in a root fileset. Recovery.
Messages FSCK Inconsistency and Error Messages 222 WARNING 222 - Catalog will be converted from up-level format Cause. A catalog at a higher version than the current version is being repaired (for example, a catalog in D40 format is being repaired by a D30.02 release version update, or RVU, copy of the FSCK utility). Effect. A new catalog is produced that is suitable for mounting by an OSS name server from the same software RVU as the FSCK utility that issued the message. Recovery.
Messages FSCK Inconsistency and Error Messages Recovery. If the REPAIR ALL option was not specified in the Subsystem Control Facility (SCF) DIAGNOSE FILESET command, the catalog is still in an inconsistent state. Reissue the command with the REPAIR ALL option. If the REPAIR ALL option was specified in the command, this is an informational message only; no corrective action is needed. 302 ERROR 302 - Invalid or Corrupt PXCKSTAT File Cause.
Messages FSCK Inconsistency and Error Messages 304 WARNING 304 - filename Purged filename indicates the affected file. Cause. The indicated Guardian disk file is either an empty orphan ZYQ file or a file belonging to an incomplete catalog left behind when an earlier execution of FSCK was interrupted. Effect. The indicated file is purged and FSCK proceeds with its checking. Recovery. Informational message only; no corrective action is needed.
Messages FSCK Inconsistency and Error Messages 307 ERROR 307 - Can’t Create SQLCAT Process - PROCESS_CREATE_ Error error:detail error:detail is the error and error detail returned by the Guardian PROCESS_CREATE_ procedure. For further information, see the Guardian Procedure Errors and Messages Manual. Cause. The FSCK utility detected a file that contains SQL/MP data but failed to start the SQLCAT process needed to delete the file. The failure might be caused by a temporary shortage of system resources.
Messages • OSS Message-Queue Server Messages Perform the following steps: 1. Use the Subsystem Control Facility (SCF) START FILESET command to remount the fileset containing the affected file. 2. Use the File Utility Program (FUP) INFO, DETAIL command or the OSS pname utility to determine the pathname of the affected file. 3. Have a user with update authority for the SQL catalog delete the file using the OSS rm utility. 4. Use the SCF STOP FILESET command on the affected fileset. 5.
Messages OSS Message-Queue Server Messages Cause. The OSS message-queue server has switched processors in response to a hardware or software failure. Effect. The OSS message-queue server now runs on what was previously its backup processor. Recovery. Informational message only; no corrective action is needed. 101 server: COMMAND LINE PARAMETERS INVALID: list server is the process name of the OSS message-queue server. list is the list of invalid parameters. Cause.
Messages • OSS Message-Queue Server Messages Supporting documentation such as Event Management Service (EMS) logs, trace files, and a processor dump, if applicable If your local operating procedures require contacting the Global Customer Support Center (GCSC), supply your system number and the numbers and versions of all related products as well. 103 server: Backup Process created in cpu nn server is the process name of the OSS message-queue server.
Messages OSS Monitor Messages Cause. The OSS message-queue server was unable to create a backup server process because of a resource shortage in the backup processor. There might be insufficient swap space on or too many processes running in the backup processor. Effect. The OSS message-queue server has no backup server process. Recovery. Correct the cause of the resource shortage. If there is a shortage of swap space, allocate additional swap space.
Messages Unnumbered Messages Unnumbered Messages Unnumbered messages can appear during startup of the OSS Monitor. Each message is prefixed by Subsystem Programmatic Interface (SPI) termination information in the following form: ABENDED: processname CPU time: interval 3: Premature process termination with fatal errors or diagnostics Termination info: code Subsystem: TANDEM.143.release text-message processname is the name of the failed process.
Messages Unnumbered Messages Invalid value specified on the command line for the AUTOSTART attribute -- must be AUTO or MANUAL Cause. The OSS Monitor detected an invalid value for the AUTOSTART attribute on its command line during startup. Effect. The OSS Monitor terminates. The existing AUTOSTART attribute value for the subsystem is unchanged. Recovery. Restart the OSS Monitor with a correct value, as indicated in the message.
Messages • Unnumbered Messages Supporting documentation such as EMS logs, trace files, and a processor dump, if applicable If your local operating procedures require contacting the Global Customer Support Center (GCSC), supply your system number and the numbers and versions of all related products as well. OSS Monitor failed in adding default record to filename file -- Error: err filename is the Guardian filename of the affected OSS configuration file.
Messages Unnumbered Messages If your local operating procedures require contacting the Global Customer Support Center (GCSC), supply your system number and the numbers and versions of all related products as well. OSS Monitor failed in creating filename file -- Error: err filename is the Guardian filename of the affected OSS configuration file. err is the Guardian file-system error number that describes the failure. Cause. The OSS Monitor could not create the indicated OSS configuration file.
Messages Unnumbered Messages OSS Monitor failed in opening filename file -- Error: err filename is the Guardian filename of the affected OSS configuration file. err is the Guardian file-system error number that describes the failure. Cause. The OSS Monitor could not open the indicated OSS configuration file. This message might indicate that the configuration file is corrupted. Effect. The OSS Monitor process terminated abnormally with the SPI error code value 17. The OSS environment is not available.
Messages Unnumbered Messages OSS Monitor failed in reading filename file -- Error: err filename is the Guardian filename of the affected OSS configuration file. err is the Guardian file-system error number that describes the failure. Cause. The OSS Monitor could not read the indicated OSS configuration file. This message might indicate that the configuration file is corrupted. Effect. The OSS Monitor process terminated abnormally with the SPI error code value 17. The OSS environment is not available.
Messages Unnumbered Messages OSS Monitor failed to get its process name Cause. The OSS Monitor could not determine its own process name. An internal error might have occurred. Effect. The OSS Monitor process terminated abnormally with the SPI error code value -7. The OSS environment is not available. Recovery. Restart the OSS Monitor with the process name $ZPMON. If the problem persists, check the Event Management Service (EMS) log for related messages.
Messages Numbered Messages 4. Restart the OSS Monitor to restart the database creation process: OSSMON /NAME $ZPMON, TERM $ZHOME, NOWAIT/ Process name of OSS Monitor must be $ZPMON Cause. The OSS Monitor was started with the name shown as processname in the SPI termination information for this message. Effect. The OSS Monitor process terminated abnormally with the SPI error code value -33. The OSS environment is not available. Recovery.
Messages Numbered Messages 1 OSS E00001 Internal error Cause. The OSS Monitor has detected a serious internal error during the execution of a Subsystem Control Facility (SCF) command. Effect. The OSS Monitor stops processing the command. Recovery. Check the Event Management Service (EMS) log for related messages.
Messages Numbered Messages ° ° ° Increase the execution priority of the OSS name server process. Start the OSS name server on a less-busy processor. Increase the cache size for the OSS name server and restart the server. 3 OSS E00003 Failed to start process -- error err, error-detail process is the process name of the OSS subsystem process that failed to start. err is the error number returned by the underlying call to the Guardian PROCESS_CREATE_ procedure.
Messages Numbered Messages Cause. A Subsystem Control Facility (SCF) ADD or ALTER command contained a parameter value that was invalid. Effect. The command is not processed. Recovery. Reissue the command with a valid value. 6 OSS E00006 No attributes have been specified for this command. Cause. A Subsystem Control Facility (SCF) command contained a parameter without a required value or was missing a required parameter. Effect. The command is not processed. Recovery.
Messages Numbered Messages 9 OSS E00009 Failed to start fileset filesetname filesetname identifies the affected fileset. Cause. A Subsystem Control Facility (SCF) START FILESET command failed. Either another fileset is mounted on the mount point you want to use for the indicated fileset, or the fileset containing the mount point you want to use is not started. Effect.
Messages Numbered Messages 3. Enter an SCF DIAGNOSE FILESET command for the indicated fileset. 4.
Messages Numbered Messages Cause. A Subsystem Control Facility (SCF) START FILESET command was issued for a fileset whose mount point does not represent a valid OSS pathname. (The pathname does not exist or is not a directory.) Effect. The fileset is not started. Recovery. Perform the following steps: 1. Use the SCF INFO FILESET command to check the spelling of the mount point pathname. 2. Use the OSS shell ls command to verify that an OSS directory exists to serve the mount point. 3.
Messages Numbered Messages 17 OSS E00017 Unable to access configuration file filename -error err filename is the Guardian filename of the OSS configuration file that could not be opened, read, or written. err is the Guardian file-system error number returned by the input or output operation. Cause. The OSS Monitor could not open, read from, or write to a configuration file. Effect. No processing can be performed on the objects in the affected configuration file.
Messages Numbered Messages Effect. The command is not processed. Recovery. A serious problem exists. Perform the following actions: 1. If possible, check and correct the configuration of the fileset using the Subsystem Control Facility (SCF) INFO FILESET and ALTER FILESET commands. Reassigning an existing attribute to the fileset might fix the configuration record problem. 2. Restart the fileset. If the problem persists, check the Event Management Service (EMS) log for related messages.
Messages Numbered Messages 22 OSS E00022 Invalid value for attribute attribute-name attribute-name is the name of the affected attribute. Cause. A Subsystem Control Facility (SCF) ADD or ALTER command failed because a specified attribute has a value that is invalid. Effect. The command is not processed. Recovery. Reissue the command using a valid value for the attribute. 23 OSS E00023 Name Server server rejected the request server is the server name of the affected OSS name server. Cause.
Messages Numbered Messages 24 OSS E00024 Name Server server gave unexpected response to OSS Monitor: status. server is the server name of the affected OSS name server. status is the status value returned to the OSS Monitor. Cause. An OSS name server returned status information that is not recognized by the OSS Monitor. An OSS name server process might be malfunctioning. Effect. The command cannot be processed. Recovery.
Messages • • • • Numbered Messages A copy of the OSS fileset catalog files Description of the problem and accompanying symptoms Details from the message or messages generated Supporting documentation such as Event Management Service (EMS) logs, trace files, and a processor dump, if applicable If your local operating procedures require contacting the Global Customer Support Center (GCSC), supply your system number and the numbers and versions of all related products as well.
Messages Numbered Messages 28 OSS E00028 Failed in moving catalog files Cause. A Subsystem Control Facility (SCF) ALTER FILESET command was issued to change the catalog volume of a fileset, but the catalog files could not be moved to the specified volume.
Messages Numbered Messages b. Use the SCF INFO SUBSYS command to check the disk volume being used for the ZOSSVOL attribute. c. If the ZOSSVOL attribute does not specify the correct disk volume, use the SCF ALTER SUBSYS command to correct the problem. 3. Check the security permissions of the CVT and FUP utilities to ensure that the OSS Monitor can execute them. Correct them if necessary. Reissue the ALTER FILESET command to change the CATALOG value. If the problem persists: 1.
Messages Numbered Messages Recovery. If the affected server is... Take this action... OSS message-queue server Use the Subsystem Control Facility (SCF) STOP SERVER command to stop the server. OSS name server Use the SCF STOP FILESET command for all filesets managed by the server. OSS sockets local server 1. Stop all applications with open AF_UNIX sockets. 2. Use the SCF STOP SERVER command to stop the server. 1. Stop all applications with open sockets. 2.
Messages Numbered Messages 32 OSS E00032 Invalid combination of command options Cause. A Subsystem Control Facility (SCF) DIAGNOSE FILESET command was issued, but the command contained an invalid combination of options. This condition usually occurs when more than one of the options STATUS, REPAIR, UPGRADE, and DOWNGRADE are specified. Effect. The command is not processed. Recovery. Reissue the command with a single option or a valid combination of options. See DIAGNOSE FILESET Command on page 12-42.
Messages Numbered Messages 35 OSS E00035 Command reserved for SUPER.SUPER only Cause. A Subsystem Control Facility (SCF) command that requires the super ID was issued but you are not logged on with the super ID. Effect. The command is not processed. Recovery. Exit SCF, change your user ID to the super ID, and re-enter SCF. Reissue the command. If you are not permitted to use the super ID, contact your system administrator to have the command issued for you.
Messages Numbered Messages Effect. The command is not processed. Recovery. Reissue the command with a single specification for the attribute. 39 OSS E00039 Missing required attribute attribute attribute identifies the missing attribute. Cause. A Subsystem Control Facility (SCF) command was entered without a required attribute. Effect. The command is not processed. Recovery. Reissue the command with the required attribute specification. 44 OSS E00044 Reserved name for OSS transport agent server Cause.
Messages Numbered Messages 46 OSS E00046 Alter Name Server of active fileset is not allowed Cause. A Subsystem Control Facility (SCF) ALTER FILESET command attempted to change the OSS name server process for a fileset while the fileset is mounted. Effect. The command is not processed. Recovery. Stop the fileset and any filesets mounted on it, then reissue the command and restart all affected filesets. 47 OSS E00047 Mount point pathname must be an absolute pathname Cause.
Messages Numbered Messages 52 OSS E00052 Fileset fileset is not mounted. fileset identifies the affected fileset. Cause. A Subsystem Control Facility (SCF) CONTROL FILESET command was entered, but the fileset targeted by it is not mounted. Effect. The command is not processed. Recovery. Use a START FILESET command to start the fileset and reissue the CONTROL FILESET command. 53 OSS E00053 Name Server server is not running. server identifies the affected server process. Cause.
Messages Numbered Messages 55 OSS E00055 The MAXINODES value is lower than the number of currently inuse inodes inuseinodes for the fileset fileset. inuseinodes is the current number of inodes in use. fileset identifies the affected fileset. Cause. A Subsystem Control Facility (SCF) CONTROL FILESET command was entered, but the specified MAXINODES value is lower than the current number of inodes in use. Effect. The command is not processed. Recovery.
Messages Numbered Messages Effect. The command is not processed. Recovery. Use the ALTER SERVER command to change the BACKUPCPU attribute, and reissue the CONTROL SERVER command with the SYNC option. 58 OSS E00058 The migration of a primary or a backup Name Server server to a different processor failed. server identifies the affected server process. Cause. A Subsystem Control Facility (SCF) CONTROL SERVER command was entered, but the backup OSS name server process is not in a valid state. Effect.
Messages Numbered Messages 60 OSS W00060 The fileset is started, but MAXINODES value is changed to maxinodesvalue maxinodesvalue indicates the MAXINODES value being used. This value is also stored in the configuration database for the fileset. Cause. The specified MAXINODES value is less than 110 percent of the number of inuse inodes. This message can occur when an OSS name server tries to mount an existing fileset and the previous maximum value did not allow reasonable growth. Effect.
Messages OSS Subsystem Messages OSS Subsystem Messages The messages in this subsection are EMS messages issued by processes running under the OSS subsystem ID.
Messages OSS Subsystem Messages For a fileset, this state is one of the following: State Meaning Diagnosing The FSCK utility was running and the fileset was unmounted. Started The fileset was mounted and its files were available for use. Stopped The fileset was not mounted and its files were not available for use. Unknown The state of the fileset cannot be determined. For the FSCK utility or an OSS server, this state is one of the following: State Meaning Started The process was running.
Messages OSS Subsystem Messages Effect. The new state determines the operations that can be performed on the object. Recovery. Informational message only; no corrective action is needed unless an SCF error message has also been sent to the console. -1 server-name cpu changed from oo to nn [ cause ] server-name is the process name of the OSS sockets local server process or OSS name server process reporting this message. oo is the processor number of the old processor.
Messages OSS Subsystem Messages file is the Guardian file identifier of the saveabend file, if any is created. Cause. The indicated OSS server process detected an internal inconsistency. Effect. The OSS server process terminates. For the OSS name server, if the SAVEABEND option is in effect, a saveabend file is generated in the subvolume where the OSS name server program file resides.
Messages OSS Subsystem Messages Recovery. See the appendix on Guardian file-system errors in the Operator Messages Manual for a definition of the specified error. For more detailed information including recovery actions, see the Guardian Procedure Errors and Messages Manual for specific information on the Guardian file-system error. Stop (unmount) the corresponding fileset and repair the fileset catalog by using the Subsystem Control Facility (SCF) DIAGNOSE FILESET, REPAIR ALL command on the fileset.
Messages OSS Subsystem Messages Value Meaning (page 2 of 2) 128 Bad parent list. The parent list in a record in the PXINODE file does not match the corresponding records in the PXLINK file. 256 Link count mismatch. The catalog and disk have different values for the number of links for a ZYQ file. 512 Missing ZYQ file. A ZYQ file is missing. 1024 Missing inode. An inode referred to in a PXLOG record is not found in the PXINODE file. 2048 Catalog disk failure.
Messages OSS Subsystem Messages Cause. An internal inconsistency interrupted execution of the indicated OSS name server process. Effect. After generating this error, the OSS name server process (either the primary or backup process) terminates. If the SAVEABEND option is in effect, a saveabend file is generated in the same subvolume where the OSS name server object file resides. Recovery. Recovery action depends on the signal number indicated in the event message, as shown in this table.
Messages OSS Subsystem Messages 5 facility EMERGENCY [,PID = pid] (identity) : text facility is the name of the OSS environment facility associated with this text message by the OSS application program. The following names are supported: Facility Name Meaning AUTH The text message was generated by a shell command or utility that belongs to a user authorization system. HP software does not use this facility name. CRON The text message was generated by a clock demon process.
Messages OSS Subsystem Messages for the process that issued the message. If the program provides no identifier, syslog appears. text is the message text supplied by the OSS application program. Cause. A site-written OSS application has sent an OSS operator message. The cause of the message is application-specific. This message identifies a severe condition requiring immediate attention. Effect. The effect depends on the application. Recovery.
Messages OSS Subsystem Messages Facility Name Meaning (page 2 of 2) USER The text message was generated by a user process. This is the default facility name used by the OSS logging feature. UUCP The text message was generated by a UNIX-to-UNIX copy program (UUCP) subsystem. HP software does not use this facility name. pid is the OSS process ID for the process reporting the error. The OSS application program can suppress display of this number.
Messages OSS Subsystem Messages Facility Name Meaning (page 2 of 2) LOCAL0 LOCAL1 LOCAL2 LOCAL3 LOCAL4 LOCAL5 LOCAL6 LOCAL7 The text message was generated by the corresponding local facility. These facility names are reserved for site-defined use. LPR The text message was generated by a line printer subsystem. HP software does not use this facility name. MAIL The text message was generated by a mail subsystem. NEWS The text message was generated by a network news subsystem.
Messages OSS Subsystem Messages 8 facility ERROR [,PID = pid] (identity) : text facility is the name of the OSS environment facility associated with this text message by the OSS application program. The following names are supported: Facility Name Meaning AUTH The text message was generated by a shell command or utility that belongs to a user authorization system. HP software does not use this facility name. CRON The text message was generated by a clock demon process.
Messages OSS Subsystem Messages for the process that issued the message. If the program provides no identifier, syslog appears. For the inetd program, this identifier is inetd. text is the message text supplied by the OSS application program.
Messages OSS Subsystem Messages Recovery. For: • • A site-written OSS application, contact your application support representative for recovery information. The OSS inetd program, no recovery action is necessary. The program adapts its behavior. If the problem was caused by a configuration error, you can do the following: 1. Edit the /etc/inetd.conf file to correct the error. 2. Use the OSS kill command to send a SIGHUP signal to the inetd program so that the program rereads its configuration file.
Messages OSS Subsystem Messages Facility Name Meaning (page 2 of 2) USER The text message was generated by a user process. This is the default facility name used by the OSS logging feature. UUCP The text message was generated by a UNIX-to-UNIX copy program (UUCP) subsystem. HP software does not use this facility name. pid is the OSS process ID for the process reporting the error. The OSS application program can suppress display of this number.
Messages OSS Subsystem Messages Facility Name Meaning (page 2 of 2) LOCAL0 LOCAL1 LOCAL2 LOCAL3 LOCAL4 LOCAL5 LOCAL6 LOCAL7 The text message was generated by the corresponding local facility. These facility names are reserved for site-defined use. LPR The text message was generated by a line printer subsystem. HP software does not use this facility name. MAIL The text message was generated by a mail subsystem. NEWS The text message was generated by a network news subsystem.
Messages OSS Subsystem Messages Recovery. For: • A site-written OSS application, contact your application support representative for recovery information. • The OSS EasySetup utilities, normally no recovery is necessary. If the reported condition causes process termination and is not a message described in OSS EasySetup Utility Messages on page A-2, refer to the error messages documented in the Open System Services Installation Guide.
Messages OSS Subsystem Messages Facility Name Meaning (page 2 of 2) UUCP The text message was generated by a UNIX-to-UNIX copy program (UUCP) subsystem. HP software does not use this facility name. pid is the OSS process ID for the process reporting the error. The OSS application program can suppress display of this number. identity is an identifier provided by the issuing program to indicate the source of the message.
Messages OSS Subsystem Messages Facility Name Meaning (page 2 of 2) LOCAL0 LOCAL1 LOCAL2 LOCAL3 LOCAL4 LOCAL5 LOCAL6 LOCAL7 The text message was generated by the corresponding local facility. These facility names are reserved for site-defined use. LPR The text message was generated by a line printer subsystem. HP software does not use this facility name. MAIL The text message was generated by a mail subsystem. NEWS The text message was generated by a network news subsystem.
Messages OSS Subsystem Messages 13 server-name SQL Catalog Manager Error err on File filename server-name is the process name of the OSS name server process reporting the error. err is the text description of the error. filename is the Guardian filename of the OSS file affected by the error. Cause. The indicated OSS name server process could not purge an OSS file using the SQL catalog manager. Effect. The OSS file is not purged. Recovery. Perform the following steps: 1.
Messages OSS Subsystem Messages Cause. An SQL catalog manager process could not be created because there is a resource shortage in the primary processor. Effect. The process is not created. Recovery. Correct the resource problem by stopping unneeded processes in the OSS name server’s primary processor. For a definition of the specified error, see the appendix describing PROCESS_CREATE_ errors in the Operator Messages Manual.
Messages OSS Subsystem Messages subsequent attempt to send the startup sequence to the backup process resulted in the error shown in detail. detail is the error detail returned by the PROCESS_CREATE_ procedure call. Cause. The indicated OSS server process cannot create a backup server process. Effect. The OSS server process has no backup process. Recovery.
Messages OSS Subsystem Messages Recovery. If error 43 is reported, make more space available on the volume or designate a new volume. The volume used for the extended swap file is determined by the setting of the SWAP attribute of the =_DEFAULTS DEFINE when the OSS Monitor ($ZPMON) is started. Positive error numbers other than 43 might indicate a more serious error with the disk volume containing the extended swap file for the OSS sockets local server.
Messages OSS Subsystem Messages mount-point is the fileset’s mount point. name-server is the process name of the OSS name server that manages the fileset. error is the file-management error number of the last attempt to propagate the mountpoint pathname. Cause.
Messages OSS Subsystem Messages 22 fileset Configured FSCK outfile, filename, unavailable, falling back to diskfile filename2 fileset identifies a fileset that is being diagnosed. filename is the filename currently configured as the value of the REPORT attribute for the fileset. filename2 indicates the alternative file that FSCK will write its report output to. Cause. The configured report file is not available. This usually means that a print spooler is unavailable. Effect.
Messages OSS Subsystem Messages 24 Monitor unable to create temporary file on volume_name, error error_number volume_name identifies the disk volume where the file cannot be created. error_number indicates the Guardian error number returned by the underlying FILE_CREATE_ procedure call. For more information about this error, refer to the Guardian Procedure Errors and Messages Manual. Cause. The OSS Monitor cannot create a necessary temporary file. Effect. The OSS Monitor terminates. Recovery.
Messages OSS Subsystem Messages objectname is either the name of the fileset or the name of the OSS name server process that is the subject of the message. The object name is represented as a subdevice of the OSS Monitor process $ZPMON. attribute is the affected attribute of the object. This attribute depends on the object type.
Messages OSS Subsystem Messages Attribute Possible Values (continued) INODECACHE 128 through 500,000 cache entries LINKCACHE 128 through 500,000 cache entries SQLTIMEOUT 60 through 300 seconds newval is the attribute value of the object after alteration. This attribute value depends on the object type. The possible values are the same as described for oldval. cause is a description of the reason for the change.
Messages OSS Subsystem Messages newvollist is a list of space-separated Guardian disk volume names; each volume name starts with a $. Each listed disk volume is eligible for OSS file creation on the fileset after alteration. cause is a description of the reason for the change. This reason is either system if the OSS name server changed the attribute value or the user ID of the operator if the attribute value changed because of an SCF command entry. Cause.
Messages OSS Subsystem Messages For an OSS name server, this attribute is one of the following: BACKUPCPU CPU INODECACHE LINKCACHE SQLTIMEOUT val is an invalid attribute value of the object. A valid attribute value depends on the object type.
Messages OSS Subsystem Messages 29 Fileset fileset pool volume list containing numVols volumes is invalid (vollist) fileset is the name of the fileset that is the subject of the message. The fileset name is represented as a subdevice of the OSS Monitor process $ZPMON. numVols is the number of pool volumes. vollist is a list of space-separated Guardian disk volume names. Cause. The OSS name server process did not recognize the specified volume list in an SCF CONTROL FILESET command issued by an operator.
Messages OSS Subsystem Messages fileset is the name of the fileset that is the subject of the message. The fileset name is represented as a subdevice of the OSS Monitor process $ZPMON. fsid is the device label of the fileset, assigned by the OSS Monitor process. catalog is the fileset catalog subvolume name of the fileset. runtime is the expected total run time in minutes for the requested operation. Cause. An SCF DIAGNOSE FILESET command began running a copy of FSCK. Effect.
Messages OSS Subsystem Messages fsid is the device label of the fileset, assigned by the OSS Monitor process. catalog is the fileset catalog subvolume name of the fileset. runtime is the expected remaining run time in minutes for the requested operation. Cause. An SCF DIAGNOSE FILESET command is still running a copy of FSCK.This event message is generated only when the FSCK process estimated run time is more than five minutes. Effect.
Messages OSS Subsystem Messages fsid is the device label of the fileset, assigned by the OSS Monitor process. catalog is the fileset catalog subvolume name of the fileset. runtime is the total run time in minutes for the completed operation. minorincons is the number of minor inconsistencies encountered. seriousincons is the number of serious inconsistencies encountered. warningincons is the number of warning situations encountered. errors is the number of error situations encountered.
Messages OSSTTY Subsystem Messages newlogfile is the name of the fileset PXLOG file after FSCK completed. Cause. FSCK finished diagnosing the indicated fileset. Effect. The fileset is available to start (mount) and use. Recovery. Informational message only; no corrective action is needed. 36 Fileset fileset has numAssocVols associated volumes fileset is the name of the fileset that is the subject of the message, represented as a subdevice of the OSS Monitor process.
Messages Startup Messages Startup Messages Usage : run osstty / NAME , IN , OUT , TERM , NOWAIT / [options] {$process-name} {$process-name | EDIT-filename} {$process-name} [options] [-backupcpu ] [-access < ALL | OWNER >] [-secure ] [-[no]server] [-[no]wrap] [-[no]prefixpid] [-coll ] [-idletimeo
Messages Startup Messages Cause. The identified flags or values cannot be used in the same command. Effect. OSSTTY is not running. Recovery. Reenter the command without specifying one of the conflicting flags. OSSTTY: Invalid attribute value for stdxx_redirection_target value indicates the value that caused the diagnosed condition. stdxx_redirection_target indicates the redirection target specification (IN, OUT, or TERM) that contains an invalid object type. Cause.
Messages Startup Messages OSSTTY: Value out of range for attribute attr attr identifies the command flag that could not be processed. Cause. The value specified is not one of the allowed values or is out of range. Effect. OSSTTY is not running. Recovery. See Starting OSSTTY on page C-1 for a description of the valid values and flags for the OSSTTY command. Reenter the command with a corrected flag or value.
Messages Startup Messages Effect. OSSTTY is running without a backup copy and cannot provide the fault tolerance of a process pair. Recovery. This is an informative message only; no action is needed unless the backup process is required by your site procedures. The OSSTTY process can be stopped and restarted after the problem is corrected. OSSTTY: ****Warning**** All the three re-directional targets are TELSERV terminals, which are accessible directly. Continuing... Cause.
Messages EMS Messages Cause. The specified collector is not accessible. The process might not be running. Effect. OSSTTY is running. Event Management Service (EMS) messages are logged to the default collector process, $0. Recovery. This is an informative message only; no action is needed unless a specific nondefault collector process is required by your site procedures. The OSSTTY process can be stopped and restarted after the problem is corrected.
Messages EMS Messages Effect. The indicated process is running. Recovery. Informational message only; no corrective action is needed. 2 OSSTTY: Process started on cpu processor-number1, without backup processor-number1 indicates the number of the primary processor for the process. Cause. The user started an OSSTTY process or the OSH command was used with the –osstty option. Effect. The indicated process is running. Recovery. Informational message only; no corrective action is needed.
Messages EMS Messages 5 OSSTTY: Trace of process osstty-name started. Logging to filename. osstty-name identifies the process being traced. filename is the Guardian filename of the EDIT file used as the trace logging file. Cause. A trace of the indicated instance of OSSTTY was initiated. Effect. The level of data logged depends on the level specified when the instance of OSSTTY was started. Recovery. Informational message only; no corrective action is needed.
Messages EMS Messages Recovery. Informational message only; no corrective action is needed. 201 OSSTTY: Takeover by backup, reason reason. reason is text that describes the reason that the backup copy of the process has taken over its open terminal sessions. Cause. The primary process was stopped by an operator action, terminated abnormally, or stopped because its processor stopped. Effect. Loss or duplication of data can occur within some terminal sessions.
Messages EMS Messages 203 OSSTTY: Unable to allocate size-in-bytes bytes size-in-bytes is the number of bytes of memory that the OSSTTY process tried to allocate. Cause. The internal allocation of memory buffers failed. Effect. Data transfer might be slow or might fail because memory cannot be allocated for data structures. Recovery. Do one of the following: • • • Limit the data traffic by using application-specific mechanisms.
Messages EMS Messages 205 OSSTTY: Maximum number of users limit has reached, currently number-of-users users. number-of-users is the number of processes (open terminal sessions) currently directing their data through OSSTTY. Cause. The number of opens of the OSSTTY process has reached the design limit for OSSTTY. Effect. All new opens are rejected; new requests from existing opens are serviced. Recovery.
Messages EMS Messages 207 OSSTTY: State of Window window-name, redirecting to proc/filename changed from old-state to new-state Reason reason window-name is the name of the OSSTTY redirection window that is the subject of this message.
Messages • • EMS Messages While the state for the OSSTTY #stdout window is SUSPENDED, OSSTTY’s #stdout window accepts new requests. If $VHS remains inaccessible even after all retries have been done, the state for the OSSTTY #stdout window changes to STOPPED. No new requests are serviced and existing requests are returned with an error. Effect. The OSSTTY window might become unavailable. Recovery. If the window state is SUSPENDED, this is an informational message only; no corrective action is needed.
Messages EMS Messages Open System Services Management and Operations Guide— 527191-002 A-108
B Manually Setting Up an OSS Environment Table B-1 summarizes the procedures for configuring and starting a new OSS environment by using manually entered operator commands instead of the OSSSETUP utility. Table B-2 on page B-8 summarizes the procedures for completing a new OSS environment preconfigured by HP or created by the OSSSETUP DEFAULTS command.
Manually Setting Up an OSS Environment Table B-1. Creating a Basic OSS Environment Without Using the OSSSETUP Utility (page 2 of 7) Task Subtask Prepare for configuration (continued) Verify system resources: • • • • • • Start related processes Example See Guardian User’s Guide 64 MB memory/ processor 500 MB disk space for root fileset Running software release version update (RVU) documente d in this guide FILEINFO $SYSTEM.SYS*.
Manually Setting Up an OSS Environment Table B-1. Creating a Basic OSS Environment Without Using the OSSSETUP Utility (page 3 of 7) Task Subtask Example See Provide required configuration files Create or modify a storage-pool file for each fileset FUP DUP ZXOSSMON.OSSPOOL, ZXOSSMON.ROOTPOOL Creating a Storage Pool on page 5-6 Ensure that preconfigured databases are accessible after a node number change FUP INFO $SYS*.*.
Manually Setting Up an OSS Environment Table B-1. Creating a Basic OSS Environment Without Using the OSSSETUP Utility (page 4 of 7) Task Subtask Example See Finish configuring the OSS environment Start an OSS shell OSH osh(1) reference page in the Open System Services Shell and Utilities Reference Manual Create the general profile file /bin/cd /etc Setting Up an /etc/profile File on page 9-2 /bin/cp profile.
Manually Setting Up an OSS Environment Table B-1. Creating a Basic OSS Environment Without Using the OSSSETUP Utility (page 5 of 7) Task Subtask Example See Finish configuring the OSS environment (continued) Configure and start network services (continued) /bin/cp smplinetd.conf inetd.conf Starting a Network Services Server on page 4-35 /bin/vi inetd.conf shell stream tcp nowait root /bin/rshd /bin/vi hosts.
Manually Setting Up an OSS Environment Table B-1.
Manually Setting Up an OSS Environment Table B-1. Creating a Basic OSS Environment Without Using the OSSSETUP Utility (page 7 of 7) Task Subtask Example See Configure OSS users Add users and specify an alias for each SAFECOM Managing Users and Groups on page 8-8 ADD USER USER.ONE, 001,001, PASSWORD Secure1 ALTER USER USER.ONE, GUARDIAN DEFAULT SECURITY NUNU ALTER USER USER.ONE, GUARDIAN DEFAULT VOLUME $DATA.
Manually Setting Up an OSS Environment Table B-2. Completing a Preconfigured Basic OSS Environment (page 1 of 3) Task Subtask Example See Prepare for configuration Log on as the super ID LOGON SUPER.SUPER Guardian User’s Guide Ensure that preconfigured databases are accessible after a node number change FUP INFO $SYS*.*.
Manually Setting Up an OSS Environment Table B-2. Completing a Preconfigured Basic OSS Environment (page 2 of 3) Task Subtask Example See Finish configuring the OSS environment (continued) Create the general profile file /bin/cd /etc Setting Up an /etc/profile File on page 9-2 /bin/vi profile umask 022 #Only users have write permission on their files. set -o noclobber #Redirection can’t overwrite files. set -o trackall #Track all aliases. export MANPATH=/usr/share/man #Match PATH use.
Manually Setting Up an OSS Environment Table B-2. Completing a Preconfigured Basic OSS Environment (page 3 of 3) Task Subtask Example See Finish configuring the OSS environment (continued) Configure utilities for periodic tasks /bin/cd /usr/bin/cron Configuring the cron Process on page 2-35 /bin/cp at.deny.sample at.deny /bin/chmod 1775 * /bin/cd /var/adm/cron /bin/cp queuedefs.sample queuedefs /bin/cp .proto.sample .proto /bin/cp cron.deny.sample cron.
C OSS Management Utilities This appendix describes: • • The optional OSSTTY utility The OSS Easy Setup utilities provided in the T0585 product OSSTTY OSSTTY can run as a single-use process or as a server. OSSTTY provides an alternative to Telserv for OSS program terminal input or output using OSS standard files.
OSS Management Utilities Starting OSSTTY Command To run OSSTTY, enter this command: [RUN] OSSTTY [ [/ NAME process_name, [ IN stdin_redirection_target, [ OUT stdout_redirection_target, [ TERM stderr_redirection_target, [ NOWAIT [ -access { ALL | OWNER } [ -backupcpu processor [ -coll collector [ -help [ -idletimeout seconds [ -[no]prefixpid [ -[no]quiet [ -[no]server [ -[no]wrap [ -secure security_string ] ] ] ] ] / ] ] ] ] ] ] ] ] ] ] ] NAME $process_name specifies the Guardian process name for the OSS
OSS Management Utilities Starting OSSTTY If an EDIT file (file code 101) is specified but does not exist, the file is created with the security specified by the -secure option or the default security setting for the user. If you omit this option, the name used is the device name for the Telserv terminal from which the command is entered. TERM stderr_redirection_target specifies the Guardian process name for the process to which data for the OSS standard error file should be written.
OSS Management Utilities Starting OSSTTY -coll collector specifies the Event Management Service (EMS) collector process to receive the OSSTTY event messages described in EMS Messages on page A-100. If you specify a collector process that cannot be used, OSSTTY issues a warning message. If you omit this option or if the specified collector process is not valid, OSSTTY uses $0 as the collector process. -help requests a summary of OSSTTY command options.
OSS Management Utilities Starting OSSTTY -[no]quiet controls whether the warning and error messages described in Startup Messages on page A-96 are displayed. Suppressing message display can be useful when OSSTTY is used within a TACL macro or other script. When the -quiet option is specified, no messages are displayed. When the -noquiet option is specified, messages are displayed as necessary. If you do not specify either option, OSSTTY uses the -noquiet option.
OSS Management Utilities • • • Starting OSSTTY The OSS shell commands and utilities do not restrict the use of the standard error file to the display of error messages. As is true for many implementations of UNIX, the standard error file can be used when unbuffered output to the terminal is the desired behavior for a utility.
OSS Management Utilities Stopping OSSTTY Stopping OSSTTY To stop a copy of the OSSTTY process started through either the OSH command or from a TACL prompt, use the TACL STOP command. Use the STOP command carefully; when OSSTTY does not terminate normally, any EDIT files it has open might not be properly closed and data could be lost. To stop a copy of the OSSTTY process monitored by the persistence manager, use the SCF Kernel subsystem ABORT PROCESS $ZZKRN.#ZTTY command.
OSS Management Utilities Utility File Security Table C-1. The EasySetup Utilities (page 2 of 2) Component Description OSSTREE file and /tmp/oss.tree. ddmmmyyyy.system_name file An EDIT file and an ASCII text file, respectively, containing a list of all directories created in the OSS file system. These files are created by the OSSSETUP utility and contain the result of the following OSS shell command: find / -WNOE -WNOG -type d -print ddmmmyyyy.
OSS Management Utilities Diagnostic Messages You can stop any of the utilities by pressing the Break key. This action is acknowledged by the following message: Break or error terminated operation. task of the OSS subsystem on system-name: FAILED Utility-name terminated unexpectedly or encountered errors.
OSS Management Utilities Utility PARAMs OSS EasySetup EMS events are logged with the OSS subsystem ID and have the form: TANDEM.OSS.D30 00010 USER NOTICE (userID) : text where userID is your user ID, and text is the event message text. Diagnostic messages for nonfatal situations include a suggested response to avoid the problem.
OSS Management Utilities OSSSETUP Utility Examples 1. To issue progress messages as EMS events using a ZSPIDEF subvolume on $SYSTEM when initially configuring and starting OSS, enter: PARAM EASYSETUP^EMSVOL $SYSTEM OSSSETUP 2. To issue progress messages as EMS events using a ZSPIDEF subvolume on $SYSTEM on the remote node \NODE when initially configuring and starting OSS, enter: PARAM EASYSETUP^EMSVOL \NODE.
OSS Management Utilities OSSSETUP Utility OSSSETUP is intended to provide a quick way to initially configure an OSS environment. You can later build on the configuration OSSSETUP creates by manually configuring additional resources or altering your initial configuration. You must not use the OSSSETUP utility when an OSS configuration already exists on your node. For example, you cannot use OSSSETUP when: • • • Your system was preconfigured by HP, but OSSREMOV has not been run.
OSS Management Utilities OSSSETUP Utility Figure C-1.
OSS Management Utilities STARTOSS Utility Considerations • • • • • • • • • Beginning with the G06.17 release version update (RVU) and the G10 version of the OSS Monitor, the OSS Monitor no longer uses the $NULL process or requires the $ZSMP process for its own startup. However, not configuring or starting $NULL or $ZSMP could affect the functioning of other products or subsystems.
OSS Management Utilities • STARTOSS Utility Your OSS environment was not preconfigured by HP or configured by using OSSSETUP, but you have created an OSSINFIL file in the STARTOSS volume and subvolume and maintain your current OSS configuration in that file. Under other conditions, using the STARTOSS utility might not achieve your intention. Command To run STARTOSS, enter these commands: VOLUME $SYSTEM.
OSS Management Utilities • STOPOSS Utility Fileset mount points should be checked after all filesets are started to ensure that they have security permissions consistent with your site’s security policies. The OSS shell chmod command can be used to modify permissions as necessary. STOPOSS Utility The STOPOSS utility stops all filesets, regardless of whether they were configured in OSSINF or OSSINFIL. STOPOSS is intended for use before a system shutdown.
OSS Management Utilities • OSSREMOV Utility You can embed this command in the subsystem shutdown file as follows: == Stop the OSS subsystem, filesets, and servers: #Push #Defaults Volume $SYSTEM.ZOSSINS Run STOPOSS #Pop #Defaults The keyword Run is optional, but it is required if #PMSEARCHLIST does not include #DEFAULTS. OSSREMOV Utility The OSSREMOV utility deletes the configurations of all processes and filesets regardless of whether they are configured in the OSSINF file or OSSINFIL file.
OSS Management Utilities • OSSINF File The OSSREMOV command prompts you to determine whether it should remove the security manager server process $ZSMP and the $NULL process. These processes are often required by other products and usually should be allowed to continue running. OSSREMOV uses the Safeguard SAFECOM program to stop $ZSMP to ensure that $ZSMP does not restart itself. If your site has not licensed Safeguard, either do not use OSSREMOV or respond “no” to its prompt about stopping $ZSMP.
OSS Management Utilities OSSINFIL File Figure C-4. Example of an OSSINF File * File: OSS Monitor in-file ( Version 1.0 ) { <--MUST BE FIRST LINE OF FILE} * This is an in-file for the OSS T0585 utilities STARTOSS and OSSSETUP. The * OSS Monitor process manages all entries in this file. * * * * There is no need to refer to any name server or transport agent server. Name servers are automatically started when their corresponding filesets are started.
OSS Management Utilities • • • SERVER entries must contain the number sign (#) character. SERVER entries for #ZTAnn are not needed. SERVER entries for OSS name servers are not valid.
D Falling Back to a Previous Release Version Update You can fall back to any G-series release version update (RVU) from a current G-series RVU or to any H-series RVU from a current H-series RVU. You cannot fall back to a G-series RVU from an H-series RVU because of hardware considerations. The procedure you use for falling back depends on the RVU to which you are falling back and the method you used to install the RVU of OSS from which you are falling back: • • For RVUs after G06.
Falling Back to a Previous Release Version Update Recreating the ZPCONFIG File Figure D-1. Sample ZPCONFIG File (page 1 of 3) == File: ZPCONFIG == == This file defines the configuration of OSS filesets. == == == DIRECTIONS FOR USE: == == CAUTION: Before starting the OSS file system for the first time, == alter the contents of this file to match your system configuration. == Add or remove one record (line) for each fileset, filled in according == to the format described in this file.
Falling Back to a Previous Release Version Update Recreating the ZPCONFIG File Figure D-1. Sample ZPCONFIG File (page 2 of 3) == Storage-Pool File == == == == == == The Guardian file identifier of a storage-pool file for the fileset. The file must be in the $SYSTEM.ZXOSSMON subvolume. == *Catalog Volume == == == == == The disk volume where the OSS name server keeps the catalog files for the fileset. == Mount Point == == == == == == The pathname of the directory on which the fileset will be mounted.
Falling Back to a Previous Release Version Update Recreating the ZPCONFIG File Figure D-1. Sample ZPCONFIG File (page 3 of 3) == FAST == == == == == == == == == == == == == == The fast-create flag, which controls buffering of new file labels by the disk process. When this flag is on, the labels for new files created on this fileset are kept in memory until the disk process has no other work to perform, then are written to disk. When this flag is off, labels are written to disk immediately.
Falling Back to a Previous Release Version Update • • Falling Back to a G06.xx Release Version Update There is an increased risk of loss of files if double failure of the disk process occurs. If there is a double failure of the disk process serving a fast-create fileset, recently created files might be permanently lost. After failure of both primary and backup OSS name server processes, when the OSS Monitor remounts a fast-create fileset, the remount is rejected by the OSS name server.
Falling Back to a Previous Release Version Update ° If there are changes to the configuration records and new features are used, you need to consider the following: • • Falling Back to a G06.
Falling Back to a Previous Release Version Update Falling Back to a G05.0x Release Version Update Falling Back to a G05.0x Release Version Update To fall back to a G05.0x RVU: 1. Back up all new files that are in OSS filesets created under a G06.xx RVU. 2. Delete all those filesets: a. At the TACL prompt, enter: SCF SCF displays its prompt, a right arrow (->). b. At the SCF prompt, enter: 1. To identify the OSS Monitor process as the default object for subsequent SCF commands: ASSUME PROCESS $ZPMON 2.
Falling Back to a Previous Release Version Update Falling Back to a G0x Release Version Update Preceding G05.00 a. At the TACL prompt, enter: SCF SCF displays its prompt, a right arrow (->). b. At the SCF prompt, make the D40 OSS Monitor the default object for SCF commands by entering: ASSUME $ZPMON c. At the SCF prompt, create the log file by entering: LOG logfile ! logfile specifies the file to which commands and output are written.
Falling Back to a Previous Release Version Update Falling Back to a G0x Release Version Update Preceding G05.00 file created in Step 1c). If the ZXCONFIG file is current, enter the following at the TACL prompt: RENAME $SYSTEM.ZXOSSMON.ZXCONFIG $SYSTEM.ZXOSSMON.ZPCONFIG • If the ZXCONFIG file is not current or if it does not exist, use a text editor to recreate the current configuration in a new ZPCONFIG file as described in Recreating the ZPCONFIG File on page D-1.
Falling Back to a Previous Release Version Update Falling Back to a G0x Release Version Update Preceding G05.00 9. Enter the following SCF command for every fileset on your system: START FILESET filesetname filesetname is the name of a fileset. 10.
E Environment Limits This appendix summarizes the architectural and functional limits that apply to the Open System Services environment on NonStop servers. Limits for a specific release version update (RVU) are subject to increases as software product revisions (SPRs) occur. OSS application programs that conform to POSIX.1 conform to a set of limits defined by the POSIX standards; POSIX limits can be less than those imposed by the environment.
Environment Limits Table E-1. OSS Environment Limits (page 2 of 4) Description OSS processes per node Maximum Value Approximately 29,000 Comment Prior to RVU G06.19 (prior to SPR T9050AOU), the limit was approximately 16,000. To Check Amount Currently Used At an OSS shell prompt: ps -lA lists active and zombie processes. OSS processes per processor Approximately 4,000 The limiting factor is the number of possible process control blocks (PCBs) per processor: At a TACL prompt, use PEEK. For RVUs G06.
Environment Limits Table E-1. OSS Environment Limits (page 3 of 4) Maximum Value Comment OSS pthreads per process At least 10,000 The limiting factor is available memory. OSS sockets per processor 4,096 OSS pipes per processor 135 to 1,024, depending on message size Prior to SPR T8620ABT, the limit was 39 to 256, depending on message size. Device labels per node ZZZZZ32 48010 Certain label designators are skipped. At a TACL prompt, use the SCF INFO FILESET *, DETAIL command.
Environment Limits Table E-1. OSS Environment Limits (page 4 of 4) Description Maximum Value OSS messagequeue messages per node 16,384 OSS message queue length 65,535 bytes OSS messagequeue message length 32,000 bytes Comment To Check Amount Currently Used A smaller limit is configurable. The default depends on the maximum number of messagequeue IDs that can be cached. At an OSS shell prompt: A smaller limit is configurable. The default is 65,535.
Glossary A7CINFO file. A Distributed Systems Management/Software Configuration Manager (DSM/SCM) configuration file distributed with OSS products that contains information about the contents of all pax archive files of an OSS distribution subvolume. absolute pathname. A pathname that begins with a slash (/) character and is resolved beginning with the root directory. Contrast with relative pathname. address space. The memory locations to which a process has access. API.
Glossary DSV the software product number and rrr is the base version identifier (such as G01) or software product revision (SPR) identifier (such as AAB.) DSV. See distribution subvolume (DSV). Epoch. The period beginning January 1, 1970, at 0 hours, 0 minutes, and 0 seconds Coordinated Universal Time (UTC). See also Coordinated Universal Time (UTC). FIFO. A type of special file that is always read and written in a first-in, first-out manner. file.
Glossary Guardian Guardian. An environment available for interactive or programmatic use with the HP NonStop operating system. Processes that run in the Guardian environment usually use the Guardian system procedure calls as their application program interface; interactive users of the Guardian environment usually use the HP Tandem Advanced Command Language (TACL) or another HP product’s command interpreter. Contrast with Open System Services (OSS). inode.
Glossary pathname resolution pathname resolution. In the OSS environment, the process of associating a single file with a specified pathname. pipe. An unnamed FIFO, created programmatically by invoking the pipe() function or interactively with the shell pipe syntax character (|). A shell pipe redirects the standard output of one process to become the standard input of another process. A programmatic pipe is an interprocess communication mechanism. relative pathname.
Glossary super ID super ID. On HP NonStop S-series or NonStop NS-series systems, a privileged user who can read, write, execute, and purge all files on the system. The super ID is usually a member of a system-supervisor group. The super ID has the set of special permissions called appropriate privileges.
Glossary working directory working directory. A directory, associated with a process, that is used in pathname resolution for pathnames that do not begin with a slash (/) character.
Index A|B|C|D|E|F|G|H|I|K|L|M|N|O|P|Q|R|S|T|U|V|W|X|Z Special Characters A A7CINFO file 6-4 ABENDED: processname (OSS Monitor message) A-31 acctcom utility (UNIX) 8-4 accton utility (UNIX) 8-4 aculog file (UNIX) 8-3 ADD ALIAS command (SAFECOM) 8-12, 8-20 ADD FILESET command (SCF) 12-8/12-14 ADD GROUP command (SAFECOM) 8-9 ADD PROCESS command (SCF) 2-9 ADD SERVER command (SCF) 12-15/12-18 ADD USER command (SAFECOM) 8-9, 8-20 Administrative groups 8-9, 8-12 adm/wtmp file (UNIX) 8-3 AF_INET sockets 1-14, 4-4,
Index B AUTORESTART attribute (continued) OSS server 12-16, 12-27 AUTOSTART attribute OSS Monitor subsystem 2-2, 2-11, 12-33, 12-35 AUTOSTART PARAM 2-11 B Backup strategies 6-15/6-16 using Backup and Restore 2.
Index D Configuration files 4-6/4-18 inetd process 4-23 storage-pool files 3-9/3-11, 4-17/4-18, 5-6 ZOSSFSET file See ZOSSFSET file ZOSSPARM file See ZOSSPARM file ZOSSSERV file See ZOSSSERV file ZPCONFIG file See ZPCONFIG file Configuration files, OSS ZPCONFIG file D-1/D-5 ZPOS files D-5 ZXCONFIG file D-1, D-4 console file (UNIX) 8-1 CONTROL FILESET command (SCF) 12-35/12-37 CONTROL SERVER command (SCF) 12-37/12-39 COPYOSS macro (TACL) 4-8, 6-4, 6-7, 6-8, 8-28, D-9 Corrupt Inode, Inode=inode-number (FSCK
Index E DIAGNOSE FILESET command (SCF) 5-22/5-32, 12-42/12-45 diff command (OSS) 3-6 change detection 8-29 dircmp command (OSS) 6-27 Directories comparing 6-27 controlling the growth of 9-8 Directory graph 5-29 Directory names, characters, number of 3-1 Dirty Catalog using Fast Create; REPAIR ALL will be performed (FSCK message) A-11 discard service (inetd) 8-6 Disk Compression Program (DCOM) 9-9 Disk files, Guardian, referencing 3-8 Disk process 1-5, 1-6, 3-7 Disk Space Analysis Program/Disk Compression
Index F ERROR messages (FSCK) 3 - operation Error error-number (description) on filename A-8 300 - Catalog Subvolume Full A-23 302 - Invalid or Corrupt PXCKSTAT File A-24 303 - Catalog Inconsistent A-24 305 - Invalid { PXINODE | PXLOG | PXLINK } File A-25 306 - DP2 Cache Flush Write Error A-25 307 - Can’t Create SQLCAT Process PROCESS_CREATE_ Error A-26 308 - Unexpected SQLCAT Error error Purging File filename A-26 6 - Catalog Version not supported by this program A-9 98 - INTERNAL ERROR A-11 99 - HEAP OV
Index F Fileset fileset pool name changed from oldpoolname to newpoolname pool volume list changed from oldvollist to newvollist by cause (OSS name server message) A-88 Fileset fileset pool volume list containing numVols is invalid vollist (OSS name server message) A-91 Fileset filesetname is not mounted.
Index G FTP (continued) initial access through Guardian 8-21 user access through 7-1 ftp user name (UNIX) 8-6 ftpd process (UNIX) 8-4, 8-5 FTPUSERS file 8-10 ftpusers file (UNIX) 8-2 Full disk volume, managing 5-21 FUP INFO display, OSS file 6-3/6-4 G games user name (UNIX) 8-6 gated process (UNIX) 8-5 GB18030 locale support 9-5 gencat command (OSS) 9-5 genxlt command (OSS) 9-6 get command options 9-11 getopts command (OSS) 9-11 getstats utility (UNIX) 8-4 gname command (OSS) 6-2 GNU 8-5 Group list 8-10
Index K INFO SUBSYS command (SCF) 12-57/12-60 ingres user name (UNIX) 8-6 Initial program 8-13 assigning 8-17/8-19 through SAFECOM 8-18 through TACLCSTM 8-19 Safeguard configuration of 8-13 use of 8-13 Initial working directory 8-13 assigning 8-13/8-17 through SAFECOM 8-14/8-17 through TACLCSTM 8-17 creating 8-14 Safeguard configuration 8-13 side effects of 8-11 use of 8-13 INITIAL-DIRECTORY attribute 4-30, 4-31 Inode cache 1-11 free list 5-28 in Guardian file identifier 3-3, 6-2 invalid number 5-31 lost
Index L Korn shell 9-1 L LANG environment variable 9-5 Large files, detecting with find command 9-8 last utility (UNIX) 8-4 lastcomm utility (UNIX) 8-4 LC_ALL environment variable 9-5 LC_COLLATE environment variable 9-5 LC_CTYPE environment variable 9-5 LC_MESSAGES environment variable 9-5 LC_MONETARY environment variable 9-5 LC_NUMERIC environment variable 9-5 LC_TIME environment variable 9-5 lib/aliases file (UNIX) 8-3 Links maximum number of 1-11 too many 5-29 links ilink number 1-6 Local sockets (AF_
Index N MINOR messages (FSCK) 205 - Missing ZYQ File, Inode=inodenumber A-16 207 - OSS Name Server Failed while Fileset was Mounted A-16 208 - There are nnn Inode Numbers Unaccounted For A-17 209 - Invalid Inode Number, Inode=inode-number A-17 213 - Orphan ZYQ File - filename A-19 214 - Catalog/File Label Mismatch, Inode=inode-number A-20 215 - Invalid Timestamp, Inode=inodenumber A-20 216 - ZYQ File Conflict - filename A-21 301 - Invalid file in ZYQ Subvolume filename A-23 Missing inode (inconsistency ch
Index O nfs_portmon environment variable (UNIX) 8-5 nice command (OSS) 2-24, 2-30 nice() function 2-24 NIS 8-10 NLS 12-5 No attributes have been specified for this command. (OSS Monitor message) A-41 nobody user name (UNIX) 8-6 Node number, changing 4-11 NOE (-W flag option) 3-6 NOG (-W flag option) 3-6 Nonzero mounted flag (inconsistency checked by FSCK) 5-28 Not a Root Fileset (FSCK message) A-21 nroff (UNIX) 9-4 ntpd process (UNIX) 8-4 null file (OSS) 8-3 NULL.
Index O OSS E00016 Unable to access catalog volume volname (OSS Monitor message) A-44 OSS E00017 Unable to access configuration file filename -- error err (OSS Monitor message) A-45 OSS E00018 Configuration contains invalid data (OSS Monitor message) A-45 OSS E00019 There is no disk volume in pool filename (OSS Monitor message) A-46 OSS E00022 Invalid value for attribute attribute-name (OSS Monitor message) A-47 OSS E00023 Name Server server rejected the request (OSS Monitor message) A-47 OSS E00024 Name
Index O OSS E00059 Unable to make all the volumes in the POOL edit file eligible for file creation (OSS Monitor message) A-59 OSS EasySetup messages A-2, A-3 OSS environment access to 8-10 backing up and restoring 6-14 differences from the Guardian environment 2-1 managed from the Guardian environment 1-1 management activities 1-3 OSS Monitor interface 12-1 restarting 2-3, 2-6 stopping 2-3 OSS file caching 5-16/5-18 OSS file compression using pack 9-9 OSS file system automatic restart of 5-8/5-11 componen
Index O OSS Monitor (continued) subsystem ID 12-6 unlicensed 8-19 verifying installation 2-8, D-9 version, determining 12-86/12-88 wait for OSS name server response 2-12 OSS Monitor failed in adding converted record to filename file -- Error: err (OSS Monitor message) A-32 OSS Monitor failed in adding default record to filename file -- Error: err (OSS Monitor message) A-33 OSS Monitor failed in creating filename file - Error: err (OSS Monitor message) A-34 OSS Monitor failed in opening filename file - Err
Index P OSS sockets local server (continued) starting 4-34 stopping 4-41 OSS software requirements 12-5 OSS standard files, redirecting C-1 OSS subsystem changing configuration of 2-18 definitions of 2-6 messages A-61/A-82 obtaining information about 2-15, 4-36 OSS transport agent servers 4-4, 12-67, 12-84 OSS W00060 The fileset is started, but MAXINODES value is changed to maxinodesvalue (OSS Monitor message) A-60 OSS W00061 The fileset is started, but not all the volumes in the pool edit file are eligib
Index Q Permissions, OSS detecting changes 8-29 FUP INFO display 6-3/6-4 Persistence count 12-16, 12-27 Persistent process OSS application 2-23/2-29 OSS Monitor 2-4, 2-7, 2-8, 2-9/2-10, C-11, D-8 OSSTTY C-1 security manager 2-9, C-11 $NULL C-11 Personalities 1-1 pgp utility (UNIX) 8-4 ph utility (UNIX) 8-4 PINSTALL command (TACL) 4-8, 6-4, 6-8, 8-28 PINSTALL utility 6-5, 6-8, D-9 pname command (OSS) 6-3 portmap process 4-5, 4-39 POSIX^CONFIG^LOC PARAM 2-12 preserve file (OSS) 8-3 Previous release version
Index S rcs utility (UNIX) 8-4 rc* files (UNIX) 8-2 rc?.d files (UNIX) 8-2 rdist utility (UNIX) 8-5 Redirecting OSS standard files C-1 Release version update (RVU), falling back to previous D-1/D-10 Remote files access to 8-20 pathnames for 3-5 Remote OSS shell commands 4-5, 9-10 REMOTEPASSWORD attribute (Safeguard) 8-20 RENAME FILESET command (SCF) 12-64 rename() function 5-15 Repair is needed for corrupted fileset filesetname (OSS Monitor message) A-49 Requests, handling, server process 4-4 resolv.
Index S SCF (continued) DELETE FILESET command 12-40 DELETE SERVER command 12-41 DIAGNOSE FILESET command 5-23/5-31, 12-42/12-45 HELP command 12-2 HELP OSS command 12-2, 12-3 INFO FILESET command 12-46/12-50 INFO MON command 12-57/12-60 INFO PROCESS command 12-57/12-60 INFO SERVER command 12-51/12-56, D-8 INFO SUBSYS command 12-57/12-60 NAMES command 12-61 RENAME FILESET command 5-33 required 12-5 START FILESET command 2-6, 5-6, 12-65/12-66, D-10 START PROCESS command 2-10 STATUS FILESET command 5-12, 12-
Index T SOCKET^TRANSPORT^NAME 4-36 SOCKET_TRANSPORT_NAME 4-36 Spooler required for printer configuration 10-1 SPR T8622AAH, falling back to a previous release version update and D-5 SQL/MP 4-2 SQL/MP programs restoring 6-24 SQL/MX compiler remote invocation 4-5 START FILESET command (SCF) 2-6, 5-6, 12-65/12-66, D-10 START PROCESS command (SCF for NonStop Kernel) 2-10 START SERVER command (SCF) 12-67 STARTMODE attribute OSS application 2-23 OSS Monitor 2-9 STARTOSS utility 2-2, 2-8, 4-26, 4-28, 4-33, 4-34,
Index U TACL (continued) PINSTALL command 4-8 STATUS command 2-8, D-9 TACLLOCL file 4-32 Tape drives not in /dev directory 10-1 Tasks, scheduling of 2-34 TCP6MON 1-19 TCPIP^PROCESS^NAME 4-30, 4-33 TCPMON 1-18 telnet command (UNIX) 8-5 telnet user name (UNIX) 8-6 Telserv configuring access 7-2/7-5 configuring user login 7-2, 8-11 direct access to OSS 7-1, 7-3/7-5, 8-11 indirect access to OSS 7-1, 7-2, 8-11 logging in through an OSS program 7-4/7-5 through OSS shell 7-3/7-4 OSSTTY break key processing C-6 r
Index V umask command (OSS) 9-3 Unable to access catalog volume volname (OSS Monitor message) A-44 Unable to access configuration file filename -- error err (OSS Monitor message) A-45 Unable to create backup in cpu nn, PROCESS_CREATE_Err: pc-error, Err detail: error-detail (OSS message-queue server message) A-29 Unable to make all the volumes in the POOL edit file eligible for file creation (OSS Monitor message) A-59 Unexpected Argument*** - token (CVT message) A-6 Unexpected SQLCAT Error error Purging Fi
Index X WARNING - Variable (variable_name) needed but does not exist (OSS EasySetup message) A-2 Warning:filename - No such File (CVT message) A-3 whatis command (OSS) 6-10 whatis database files 6-1, D-10 whatis database (OSS) 6-11 wheel group name 8-6, 8-7 who user name (UNIX) 8-6 Wildcard characters 2-13 Wrong fileset type (inconsistency checked by FSCK) 5-30 X xargs utility (UNIX) 8-5 Z ZEMSTACL C-9, C-10 ZFB* file 6-6, 6-8 ZINSPOOL 4-18 ZOLDFSET 4-10 ZOSSFSET file 3-8, 4-6/4-12 automatic creation 5-
Index Special Characters $SYSTEM.SYSTEM.UNISTDH 9-6 $SYSTEM.ZRPC.RPC 4-25 $SYSTEM.ZTCPIP.HOSTS 4-24, 4-25 $SYSTEM.ZTCPIP.IPNODES 4-24 $SYSTEM.ZTCPIP.NETWORKS 4-24 $SYSTEM.ZTCPIP.PORTCONF 4-23 $SYSTEM.ZTCPIP.PROTOCOL 4-24 $SYSTEM.ZTCPIP.RESCONF 4-24, 4-25 $SYSTEM.ZTCPIP.SERVICES 4-24 $SYSTEM.ZXOSSMON 5-6 $VHS C-1 $YMIOP.
Index Special Characters /etc/shutdown utility (UNIX) 8-2, 8-4 /etc/syslog utility (UNIX) 8-2, 8-4 /etc/termcap file 8-3 /G directory 1-6, 1-7, 3-5, 4-7, 6-26 /home directory 4-7 /home/quotas file (UNIX) 8-2 /nonnative/usr/share/man directory D-10 /tmp directory 4-7, 5-3, D-5 /tmp/oss.tree.ddmmmyyyy.system_name file C-8 /usr/etc/exportfs utility 8-5 /usr/etc/showmount utility 8-5 /usr/include directory 6-7 /usr/include/unistd.h 9-6 /usr/lib/cron/at.allow file 2-37 /usr/lib/cron/at.
