Setup for the IT Service Vision Server Introduction Preliminary Checklist Section 1, Getting Up and Running The purpose of this section is to guide you through the steps that are necessary to collect representative performance data at your site, create a test IT Service Vision PDB, process the data into it, reduce the data in the PDB, and analyze and report on the data.
Cabletron SPECTRUM Appendix 3: SPECTRUM Specific Tips for Reporting Cabletron SPECTRUM Appendix 4: Defining IT Service Vision Tables for SPECTRUM Data Cabletron SPECTRUM Appendix 5: Editing Model Types for SPECTRUM Cabletron SPECTRUM Appendix 6: Mapping SPECTRUM Data Types to an IT Service Vision PDB Open Systems | OpenView and NetView OpenView and NetView Appendix 1: Recommended Metrics OpenView and NetView Appendix 2: Alternate Data-Logging Method OpenView and NetView Appendix 3: OVW-Specific Tips for Rep
Introduction This documentation contains setup instructions and information for the IT Service Vision server for Open Systems and Windows NT supplied collectors. Some of the instructions and information apply across all platforms, all collector facilities, and all data sources and collectors. Some of the instructions and information apply only to a specific platform, facility, data source, or collector. The instructions and information that differ are identified with icons.
Label Icon Collector ACC UNIX Accounting CSi SPECTRUM by Cabletron Systems, Inc. MWA Performance Collection Software and MeasureWare by Hewlett-Packard Company NV NetView for AIX by IBM OV OpenView Network Node Manager by Hewlett-Packard Company PRX PROBE/Net by Landmark Systems SNM SunNet Manager by Sun Microsystems, Inc. TKR TRAKKER by Concord Communications, Inc. Note: In this document the label OVNV and the icon represent both OpenView and NetView products.
Label NT SMF ROLM Collector NT SMF by Demand Technology Software Rolm PBX Traffic R/3 R/3 by SAP AG Web Web Server Log Data DECps Polycenter Performance Solution by Digital Equipment Corporation SNMP SNMP mib-II data vmstat UNIX vmstat Additionally, the IT Service Vision Showroom contains information on how to web-enable reports on data from any collector.
Preliminary Checklist This checklist describes the information that you may need to gather and verify before you set up IT Service Vision. This checklist also has advanced topics that you may want to make use of after setup. Suggested Reading In this document, you will see references to SAS windows, pull-down menus, and command lines. For more information on using these in IT Service Vision, see Shared Appendix 2: Navigating SAS Windows.
Because Open Systems software comes from so many vendors, the best sources of general purpose information are books. The following are recommended: Applications for Distributed Systems and Network Management . By Kornel Terplan and Jill Huntington-Lee. Van Nostrand Reinhold. 1994. ISBN 0-442-01873-8. System Performance Tuning. By Mike Loukides. O’Reilly and Associates. 1991. ISBN 0-937175-60-9. UNIX System Administrator’s Handbook, Second Edition. By Evi Nemeth, Garth Snyder, and Scott Seebass.
HP OpenView Network Node Manager User’s Guide HP OpenView Network Node Manager Administrator’s Reference HP OpenView Windows User’s Guide IBM NetView for AIX User’s Guide for Beginners IBM NetView for AIX Administrator’s Guide IBM NetView for AIX Problem Determination. Abbreviation and Terms for Cabletron SPECTRUM Caution: The SPECTRUM examples and definitions are meant as a guide to setting up SPECTRUM. The examples and definitions were created using SPECTRUM Release 4.0 and work with that release.
An attribute is the name of a variable (metric) that is defined for a model type and whose values are assigned or measured for the models of that model type. The attributes for each model type are controlled by that model type’s Management Module. For example, some of the attributes for Cabletron MRXI repeaters are Model_handle (model id), Model_type_handle (model type id), and IP_In_Discards (the number of IP layer packets discarded).
Microsoft Windows NT® Workstation Resource Kit. By Microsoft Press. 1996. ISBN 1-57231-343-9. See Part III, Optimizing Windows NT Workstation. See the white paper written by Demand Technology: Windows NT: Can it be tuned on the World Wide Web at http://www.demandtech.com/ See the white paper written by Compaq: Performance Tuning Your Windows NT Web Server on the World Wide Web at http://www.compaq.com/support/techpubs/whitepapers/457a1096.html.
IBM NetView for AIX Administrator’s Guide IBM NetView for AIX Problem Determination.
Section 1, Getting Up and Running The purpose of this section is to guide you through the steps that are necessary to collect representative performance data at your site, create a test IT Service Vision PDB, process the data into it, reduce the data in the PDB, and analyze and report on the data.
Section 1, Task 1: Start the IT Service Vision Server Software This task walks you through the basic site-level customization of the IT Service Vision server software. (Customization of the the IT Service Vision client software is covered in a later task.) Prerequisites 1. You know the work shift schedule and holiday schedule at your site. You need a schedule of the work shifts that are important for your performance reporting. You can define as many as six shifts with the server’s interactive interface.
or !cproot/misc/itsv -sasroot sasroot b. To start SAS and then start IT Service Vision, issue these commands: i. Invoke SAS by issuing this command at your shell prompt sas -dms & (or the appropriate site command if your site customized the command for starting SAS) ii. From within SAS, do one of the following Follow this path Globals -> Options -> Command -> type CPE -> OK Or, in the body of the PROGRAM EDITOR window, type %cpstart(); and then follow this path Locals -> Submit iii.
Or, in the body of the PROGRAM EDITOR window, type %cpstart(); and then follow this path Locals -> Submit iii. After IT Service Vision server interface displays the IT Service Vision splash screen. Select OK to Continue. Now, IT Service Vision displays the IT Service Vision main window. If you installed ITSV2 in a temporary directory (that is, not in the same directory that the SAS System is in), you need to specify the location of IT Service Vision when you start it.
the PDB, and write access to SITELIB: %CPSTART( pdb=’/tmp/pdb-test’, siteacc=write ); This statement invokes the IT Service Vision server with a pdb named pdb-test as the active PDB, write access to the PDB, and read access to SITELIB: %CPSTART( pdb=’/tmp/pdb-test’, access=write ); This statement invokes the IT Service Vision server with a pdb named pdb-test as the active PDB and write access to the PDB and write access to SITELIB %CPSTART( pdb=’/tmp/pdb-test’, access=write, siteacc=write ); Then, from your
Modify the default list so that it correctly represents the holiday list at your site. To change the status or description of a date already on the list, select the date and then right mouse click and Edit. Change the status and/or description, and then select OK. To add a date to the list, select Actions and Add Holiday. Fill in the date, select Active, fill in the description, and select OK. To delete a date from the list, select the date, select right mouse click and Delete, and confirm with OK.
Section 1, Task 2: Start Your Data Collection Software This task walks you through checking the installation of your logging and collector software and then starting the logging and collection. Prerequisites In this document, the host on which your collector software is installed is called the collector host or the management host. You can use your workstation as the collector/management host. However, typically the collector/management host is a different machine from your workstation. 1.
With SPECTRUM Release 4.0, the recommended method for starting the server is by using the SPECTRUM Control Panel. Using the SPECTRUM Control Panel insures that the environment is properly set up and that the supporting daemons, DAS and ArchMgr, are running. For versions prior to 4.0, please refer to your SPECTRUM System Administrator’s Guide for the proper method for starting SpectroSERVER. i. Check that the SPECTRUM processd daemon is started.
Note: If, instead of the eye window and View window, SPECTRUM displays a window with a message similar to this one Error: No SpectroSERVER to talk to! then SpectroSERVER is not running on the selected host. Select Close on this window. Repeat this action and select the correct host name (the one that you selected for SpectroSERVER). In this document we suggest you export and examine the data one model type at a time. Select a model type.
When an icon representing a HubCSIMRXi is visible, follow this path View -> Icon Subviews -> Model Information to select the icon subview that controls polling and logging of attributes for the HubCSIMRXi devices. For model types other than HubCSIMRXi, use the Management Module Guide to determine which icon subview controls polling and logging of attributes for your device. Set the Poll Interval field to 60 to collect data every minute.
SASROOT=/local/sas export SASROOT iv. Save the edited version and exit from the editor. Note: If you want to know more about dtxscript, see the document SPECTRUM Data Export User’s Guide. f. Export the data using SDE and the export script. SDE uses Export Definition Descriptions, which are templates that control the export process. Because this is your first time using SDE, you will need to create one of these descriptions. i. Start SDE.
name for all data in your statistics data set. iii. Save the Export Definition Description. Select: File -> Save... In the Selection field, type a file name such as exp_def_desc_1 (SDE automatically appends the .sde suffix to the file name ), and then press ENTER or RETURN. In exp_def_desc_1.sde, SDE saves the description of what data to export. g. Allow SPECTRUM some time to collect data.
tail -f SPECTRUM_home /export.output/exp_def_desc_1.log where exp_def_desc_1 is the name of the Export Definition Description that you created in this action. Note: Tail the .log file not the .sde file. When SPECTRUM displays a message in the UNIX window that indicates the export of the models data and statistics data is complete, press CTRL-C in the UNIX window to interrupt the tail command. j. Close SDE.
Use FORMAT BINARY instead of FORMAT ASCII. IT Service Vision expects the exported data to be in binary format. Use LAYOUT SINGLE instead of LAYOUT multiple. LAYOUT MULTIPLE data typically contains many placeholder records (ones with zero or negative counts) that are difficult to detect and remove from your PDB and which serve no purpose for long term analysis. Do not request CONFIGURATION data. CONFIGURATION data is not supported by IT Service Vision.
object manager name snmpCollect behavior OVs_WELL_BEHAVED state RUNNING PID 28641 last message Initialization complete. exit status - ii. Check that the value for state is RUNNING, for each of the four background processes.
NetView Tools -> Data Collection & Thresholds: SNMP OpenView Options -> Data Collection & Thresholds: SNMP... The MIB (Management Information Base) data collection window displays. The top third of this window lists which MIB objects, if any, are already configured for data collection. Some recommended metrics for your test PDB are listed in OpenView and NetView Appendix 1: Recommended Metrics. To add the selected metrics to your PDB, follow these steps: i. Select Metrics Select Add...
Select Suspend and then Resume. iii. Verify that data are being collected. You can execute the following command periodically: snmpColDump databases/snmpCollect/ifIndex. n where n is the number of the interface. For instance, if n is 1, this command dumps all the ifIndex raw data that were collected until now for interface number l. f. Allow the management software some time to collect data. For best results at this stage, allow the management software to collect data for about two days.
Set up TRAKKER to log the data of interest to disk at some regular interval (usually 15-60 minutes). Although in theory TRAKKER will continue to log data until your disk space is totally consumed, we recommend that the log file be moved out of the path of TRAKKER, processed into your PDB, and then archived to tape fairly frequently (daily).
Use LAYOUT SINGLE instead of LAYOUT multiple. LAYOUT MULTIPLE data typically contains many placeholder records (ones with zero or negative counts) that are difficult to detect and remove from your PDB and which serve no purpose for long term analysis. Do not request CONFIGURATION data CONFIGURATION data is not supported by IT Service Vision. When using CSPROCES to read MeasureWare data, you may get the message "Variable XXXXXXX is uninitialized" in your initial run with a table.
ii. Check that the value for state is RUNNING, for each of the four background processes. If the state value for one or more of the four services is not RUNNING, issue the following command from a DOS command window: ovstart b. Start the SNMP management software graphical user interface. Start->Programs->HP OpenView->Network Node Manager Perform all the actions in this task from within the SNMP management application. c. Select all the nodes from which you want to collect data. i.
Select Add From Map. The nodes that you selected in the previous action are added for this metric Select Collection Mode and then Store, No Thresholds. Type 15m (for 15 minutes) in the Polling Interval field. Select Instances and All. Select OK for your addition to take effect. The MIB object now displays in the list of configured MIB objects (metrics) in the top third of the screen. ii. Repeat this sequence for the remaining metrics.
Section 1, Task 3: Create a Test PDB and Process, Reduce, and Report on Your Data Create a test PDB and add one or more tables to it or create tables in it. Process your data into the table(s), reduce the data, generate reports on the data, and view the reports. Prerequisites 1. You have logged data for approximately six hours to one day, depending on the amount of data that you want to collect. 2. You have determined which day is considered the start of the week at your site.
name derived from the collector name). In the Wizard, select Next to Continue. b. Add or create table and variable definitions. i. Use the IT Service Vision user interface to add the tables and variables. Select Select Tables. IT Service Vision displays the Select Tables window. ii. Select the table(s) based on your collector. From the collector list, select your collector. In the table list select your table(s).
Vision tables from schema files appear in SunNet Manager and Enterprise Manager Appendix 2: Defining Tables from Your Own Schema Files. The table definition process adds the tables directly to your PDB (so you do not select the tables from the Select Tables list). For other data sources, select the tables in the Select Tables list that correspond to the data in which you are interested by matching the table description displayed to your collector vendor’s description of the data.
The reduce data wizard can be started in two ways: from the main windows, select Reduce Data - Wizard at the end of the process data wizard, continue with the reduce data wizard. Follow the directions in the reduce data wizard. At the end of the reduce data wizard, you have an opportunity to save the source code for this invocation of the reduce macro. Save the source code; you will use it in Section 2. It will be necessary for the batch processing examples in this document.
preparation (data manipulation) prior to running it. (This is listed after the heading PreReqs in the Explanation.) In a later step you will run the reports that do not require setup. Close the window. ii. View sample output from the report. Follow this path for the same report definition: Locals -> Show sample output Close the window. Select and run each report definition that does not require setup.
Then follow this path: select report to view or edit -> File -> Open -> Browse or Edit
Section 2, Making the Transition to Production Section 1 got you started with data collection, a test PDB, and some simple report definitions. This section describes how to finish customizing the data collection, PDB, and report definitions to the specific requirements at your site. It also describes how to put the customized versions into production.
Section 2, Task 1: Customize and Verify Your Test PDB This task walks you through planning which variables to keep in your production PDB. It also walks you through implementing decisions in the test PDB, checking the test PDB, and adjusting the test PDB until you are satisfied with your choices. In the task after this one, you will make a production PDB from the customized test PDB. Prerequisites 1. You have a printed copy of the table definition(s).
i. Create a file containing these statements: * Printing a table definition in the SAS log; %CPSTART( pdb=pdbname, mode=batch ); %CPCAT; cards4; print table name=_ALL_; /* The four semicolons must be in column one */ ;;;; %CPCAT( cat=work.temp.prtab.source ); %CPDDUTL( entrynam=work.temp.prtab.source, list=yes ); where pdbname is the name of your PDB. For more information on the PRINT TABLE control statement, see the %CPDDUTL control statements in the Macro Reference documentation for IT Service Vision.
Each of the BY and CLASS variables must have a Kept status of Yes. Note: These lists apply to a table definition and are used to process and reduce data. This BY variables list is not related to the BY variables that you can select for a report definition. More about frequently used variables: MACHINE is the host name of the node on which the metrics were created. In the process step, the value of MACHINE is obtained from the value of IPADRES. DATETIME is the timestamp on the observation.
Index variables are variables whose values you want to index at one or more level. The purpose of an index is to speed up the generation of reports when you run a report definition that has one or more of the indexed variables. d. Review the other variables in your table to check whether their Kept status and statistics to be calculated are appropriate for your site.
production jobs when the data will not be used again. There is one caution: if you have a failure (for instance, a power outage) during the reduce step and any age limits at detail level are zero and any age limits at non-detail levels are non-zero, make sure that you re-submit the reduce step before you go on; otherwise, the next day’s production job will remove the detail data before the data have an opportunity to be summarized in the non-detail levels. ii. Check the Age Limits at reduction levels.
Follow this path from the IT Service Vision main window: Administration -> Manage Tables IT Service Vision displays a list of the tables in the active PDB. b. Delete the data. Follow this path: Manage Tables-> select a table -> right button to bring up popup menu -> select ’Delete Table’ -> Delete data only c. Return to main window. File -> End 2. As necessary, make changes to your PDB’s work shift schedule, holiday list, and start of week.
If you decide in the task before this one, to change the SPECTRUM attributes for the model/table in your PDB, now is the time to add variables that correspond to these metrics. See Cabletron Spectrum Appendix 6: Mapping SPECTRUM Data Types to an IT Service Vision PDB in this document for information on adding new variables to a table. To add new variable(s) in the View Variables window, follow this path: select variables -> close then follow the information under [help] on each tab.
Select or de-select statistics to correspond to your choices. (Only the selected statistics will be calculated.) Then select OK. When you finish editing the statistics for the variables whose Kept status is Yes, return to the table definition by selecting OK. g. Modify the BY and Class variables lists, ID variables lists, and index variables lists, if you require values different from the default settings. The BY variables list specifies the sort order and grouping of data at the detail level.
After you finish customizing a table and its variables, on the table definition/properties window select OK. It can take a long time for the data dictionary to be updated and the view or views to be recreated, perhaps several minutes. If you have time to do this now, select OK; otherwise, select Cancel. (The next action describes what to do later if you select Cancel now.) Then IT Service Vision returns you to the list of tables in your PDB.
Note: This Task describes adding a metric by clicking down the OID tree. It is somewhat faster to type the metric’s OID in the MIB Object Id field and press ENTER or RETURN. To obtain the OIDs for the metrics that you want to change, refer to your copy of the complete table definition (see Prerequisites for this task). 6. Process and reduce a day’s data.
Another consideration of using ifSpeed and ifUtilization is determining whether an interface is half-duplex or full-duplex. In the calculation of ifUtilization that is supplied with IT Service Vision, all interfaces are assumed to be half-duplex. Interface utilization is thus calculated using the sum of input and output octets ifutil = (ifiocts+ifoocts)*8 / ifspeed If you use concentrators on your lines or if the interface is full-duplex, utilization may sometimes be reported as exceeding 100%.
f. Close the FSVIEW window. The SAS informat CSFDX is used by the IT Service Vision process step when it calculates ifUtilization, so for the utilization calculation to be done properly you must run the above statements before you run the process step. If you add or change full-duplex interfaces, you will need to run these statements again. The new information will be utilized the next time that you run the process step. b. Check the SAS log for error messages. c.
Section 2, Task 2: Create Your Production PDB You now have the test PDB customized the way that you want it. This task will walk you through creating your production PDB and copying the structure and settings and data of your test PDB to your production PDB. Prerequisites 1. Your test PDB is the active PDB. 2. You have write access to your test PDB. 3. You have chosen a name for the PDB that this task will produce. In this task you will create a permanent PDB for your collector’s data.
From the main window in IT Service Vision, follow this path: Administration -> Create PDB - Wizard Provide the name of the new PDB. Then select Next in every window and Finish in the wizard’s final window. b. Copy your test PDB’s structure, settings, and data to your production PDB.
Section 2, Task 3: Customize Your Report Definitions In this task, you will customize the report definitions that you used in Section 1. Prerequisites 1. Your production PDB is the active PDB. Administration -> manage PDBs -> select the production PDB -> Activate -> WRITE -> OK -> OK 2. You have decided on storage locations for your customized report definitions. Report definitions exist in two forms, interactive and batch/background. The interactive form is ready to run in the IT Service Vision GUI.
Note: You can also add footnotes. c. Exit the Set Titles and Footnotes window. Select OK. 3. Save the customized report definition in interactive form. From the Manage Report Definition window, follow this path File -> Save Report Definition IT Service Vision displays the Save Report Definition window. If the original report definition was supplied, the Name and Description are copied from the supplied report definition.
5. Repeat these actions for the other report definitions that you used in Section 1.
Section 2, Task 4: Set Up Your Production Job This task walks you through setting up and scheduling the batch or background job that will automatically process, reduce, and report on your site’s data. Prerequisites 1. You understand how to use the job scheduling system at your site. Refer to the documentation for your task scheduler for more information. 2. You understand the macros for return code checking. See the Macro Reference documentation for more information on return code checking. 3.
Note: Any mention of the pr_script later in this document refers to the processing and reducing script for your collector. b. the batch/background files that you generated in Section 1, Task 3: Create a Test PDB and Process, Reduce, and Report on Data and revised in Section 2, Task 3: Customize Your Report Definitions. i. Edit the start, process, and reduce code into a single file in the following order: The code that starts IT Service Vision and processes the data into the PDB.
the code that reduces the data in the PDB. This combined file is your collector’s version of the pr_script file. ii. Edit the code for generating the reports into a single file in any order. This combined file is your collector’s version of the reporting macros that are included in the IT Service Vision prototype file named csdaily. Note that csdaily also includes other statements that you will need. Actions 1. Create the production job(s).
b. Edit your crontab file to execute your scripts that process, reduce and report on the data. i. Retrieve your current cron file. You will be adding jobs to your crontab file. Retrieve your current crontab file with the following command: crontab -l > my_crontab ii. Edit my_crontab file.
Section 2, Task 5: Prepare for Client Use of PDB(s) This tasks readies the IT Service Vision server for use with the IT Service Vision client. You can skip this task if your site does not plan to use the IT Service Vision client. Note: The IT Service Vision client can use remote access to run "directly" on the data in the PDB(s). Or the IT Service Vision client can run on a downloadable subset of data from the PDB(s). You can prepare for either or both of these options.
For more information on all of the fields in a remote profile, follow this path: From the main window, select Administration -> Manage PDBs -> File -> Include PDB -> click on the rightward arrow that is associated with the Remote Server Profile field -> UNIX or MVS-> OK-> Help Note: Because IT Service Vision clients usually have network access to the server if the server is on Windows NT, the clients rarely need to use a remote profile to access a PDB on a Windows NT server. 2.
modify your copy of the autoexec.sas file to remove references to the applications you did not copy, change your working directory to the directory containing the Desktop Reporter files, and start SAS and the Desktop Reporter in the normal way. 2.REMOVE UNNEEDED SUPPLIED REPORTS To remove unneeded reports, you will need to have licensed either IT Service Vision or SAS/AF. Start the SAS system and enter the EIS command in the command area of the tool bar.
Sales. STRUCTURE OF OPEN SYSTEMS DESKTOP REPORTER The following is a concise representation of the menu structure of the Open Systems DeskTop Reporter. This list can be used to identify the names of components that produce a report so as to modify or remove references to the component.
o.download.listmenu o.signunix.af o.signunix.scl i.signon.program o.downdata.af o.downdata.scl i.download.scl i.downprof.scl i.downfrom.program i.signoff.af i.signoff.scl i.downunix.af - o.osyshelp.listmenu i.using.af i.downunix.af o.modify.af - Download Open Systems Data Signon to UNIX Setup download from UNIX Signon to UNIX Download Data from UNIX Set applid=OSYS for download Download data from remote host Create SASUSER.
Section 2, Task 6: Wrap Up This task walks you through the steps necessary to make the production PDB and the production reports available to IT Service Vision users. It also walks you through other steps that complete the transition into production mode. Prerequisites 1. You know the complete names of the production PDB, SITELIB, PGMLIB, and platform-specific structures. a.
Administrator User UNIX WNT read misc/cpe -- Administrator User cpe\sasmisc read -- Note: For a summary of which activities require which access rights to which structures, see Shared Appendix 3: Actions and Access Rights Required. 2. Add items for recurrent tasks to your reminder file. Update your site’s holiday list in SITELIB once a year. For details, see Section 1, Task 1: Start the IT Service Vision Server Software. Update your site’s work shift grid at seasonal shift changes.
Section 3, Advanced Setup
Section 3: Advanced Setup You may want to consider: 1. Deleting one or more demonstration PDBs. At some sites, copies of the demonstration PDBs are used for practice or testing; for instance, testing what a new report would look like when the production PDB does not have enough data to generate a report that looks good. At some sites a demonstration PDB is used when the active PDB must be some other PDB than the one that is to be used. (Some operations cannot run on the active PDB.
Note For UNIX and WNT, do not select tables to add. Do request that the new empty PDB become the active PDB. b. Change to another PDB as active PDB. To list the known PDBs, follow this path from the main menu: Administration -> Manage PDBs IT Service Vision displays your list of known PDBs with the name of the active PDB in the header. Switch to any other PDB than the PDB that will contain the combination of tables (for instance, switch to a demonstration PDB) as the active PDB.
Consider creating new report definitions. Decide if you want to use any of the other HN2 tables and, if so, which other variables you want to use from those tables. Consider collecting SNMP data using the other MIBs for which IT Service Vision supplies tables. Consider collecting SNMP data using MIBs for which IT Service Vision does not supply tables. Consider creating new report definitions In this example, you will create a new, custom report definition..
b. To create a new report definition, select File -> New Report Definition. c. To select the table whose data is to be used in the report definition, select the arrow to the right of the Select Table field, and then follow this path, starting at All: CSi All -> DAY.CSIMRX -> OK OVNV All -> DAY.HN2NIX -> OK OV All -> DAY.HN2NIX -> OK d.
You can also make any number of custom report definitions using Manage Report Definitions on the Reporting tab. Using UDAY and UWEEK is not necessary. You can use any variables in the specified table and any other features available in the specified type of report definition. Note: If you make custom report definitions, remember to add them to the appropriate batch/background job file.For details, see Section 2, Task 3: Customize Your Report Definitions.
To construct IT Service Vision tables that correspond to the MIBs for which IT Service Vision does not supply tables, you need to run the IT Service Vision mib2dict program on the MIB. For instructions on running the program, see Open Systems and Windows NT Appendix 4: MIB to Dictionary Compiler.
Platform - Specific Appendices
Open Systems and Windows NT Appendix 1: Tables and Variables Naming Convention SUPPLIED TABLE NAMING CONVENTION IT Service Vision supplied tables are named according to the following scheme: cmmnnn || | || | || +-----Name: Abbreviation for the group of metrics in || this table || || For example, || ICM - nb-mibII.icmp || NIX - nb-mibII.
For example, the description field for table HN2ICM begins nb-mibII.icmp and for table HRMATM begins RFC1271-MIB.alarmTable. USER-DEFINED TABLE NAMING CONVENTION Begin the names of user-defined tables with the character U. must be exactly six alphanumeric characters.
Open Systems and Windows NT Appendix 2: Tips for Reports A Collection of Tips Tips for Subsetting Data Subsetting via DATETIME Subsetting by a Simple Local Where Expression Subsetting by a Simple Global Where Expression A Collection of Tips Here are some tips that may be useful to you as you generate reports with the IT Service Vision GUI. If running your report definition does not seem to produce any results, be sure that you are looking for the report in the correct window.
select the table -> click right mouse button -> Properties Check that the Kept status of the table is set to Yes. Then select Variables and check that the Kept status of the appropriate variables (select a variable -> click right mouse button -> Properties) is set to Yes, and the appropriate levels are set to non-zero durations. For the non-detail levels, check that the appropriate statistics are selected.
IT Service Vision looks in the PDB’s data dictionary to find the earliest and latest DATETIME values at the level that is specified on the main report definition window. If you want to use less than that range, you can type over the values in the Begin and End fields. Note: The need to reset may occur when you are designing or modifying report definitions and switch from one table to another or one level to another.
Open Systems and Windows NT Appendix 3: Printing Graphs Introduction Printing Graphs from a Batch Job Prerequisites Actions Introduction This appendix steps you through the task of customizing the printing of graphs interactively and in background.
1. Ensure that you have a GOPTIONS statement that includes a DEVICE= parameter, and set the DEVICE= parameter to the same device driver that you selected above. (You do not need to specify the TARGETDEVICE= parameter.) For example, goptions device=hplj4si ; 2. Then specify settings by means of OUT xxxx parameters on the IT Service Vision reporting macros. For example, to route the graphs to a printer/plotter, you would use these parameters on each of the reporting macros: outmode=catalog outloc=work.
Open Systems and Windows NT Appendix 4: MIB to Dictionary Compiler Introduction Creating Table and Variable Definitions from a MIB Prerequisites Actions Introduction In the master data dictionary, IT Service Vision supplies table and variable definitions for the data from many collectors. For those data, you can use the supplied table and variable definitions (after adding them from the master data dictionary to your PDB).
For the purposes of this exercise, we will use rfc1213.mib, which is a file containing the definition of MIB-II. Assuming you have installed SAS and the IT Service Vision server in /usr/local/sas, you can find a copy of rfc1213.mib in /usr/local/sas/saspgm/cpe/mibs/. Actions 1. Compile the MIB. In this action we are going to compile rfc1213.mib and produce two output files: rfc1213.ddu and rfc1213.sas . The .ddu file will contain the CPDDUTL control statements needed by the %CPDDUTL macro. The .
not in the table’s BYVARS or CLASSVARS lists. In particular, do not delete MACHINE, HOUR, or SHIFT because these variables are always on the BYVARS and CLASSVARS lists. For more about the BY and CLASS variables lists, see Section 2, Task 1: Customize and Verify Your Test PDB. 2. There may be some IT Service Vision table or variable names that you want to change .
Collector - Specific Appendices
Cabletron SPECTRUM Appendix 1: Starting the processd Daemon The processd daemon is not running on the specified SpectroSERVER if you start SpectroSERVER and get a message similar to this one: SpectroSERVER: Cannot connect to the processd at this location: machine name where machine name is the SPECTRUM management host name. To start the daemon, perform the following steps on the SPECTRUM management host machine. 1. Check to see if the processd daemon is running.
Cabletron SPECTRUM Appendix 2: Alternate Data-Logging Method This is an alternate version of the task in Section 1, Task 3: Create a Test PDB and Process, Reduce and Report on Your Data. In this version of the task, you update the SPECTRUM database using the SPECTRUM Command Line Interface (CLI). Using CLI, you can enable logging of attributes for more than one model at a time. Prerequisites 1. SAS is installed. See Section 1, Task 2: Start Your Data Collection Software. 2.
Set this to the SPECTRUM base directory to which SPECTRUM_home refers. For example, if SPECTRUM is installed in /usr/spectrum, DATA_PATH=’/usr/spectrum’. DATA_DIR Set this to the directory (under the base directory) where the command line interface programs are located. Normally, DATA_DIR=’vnmsh’. POLL_INT Set this to the number of seconds between polls of the devices. A smaller number will create more network traffic. LOG_RATIO Set this to the fraction of polls that are logged.
Cabletron SPECTRUM Appendix 3: SPECTRUM Specific Tips for Reporting Using One or More Formula Variables Formula variables are variables whose values are calculated (from the values of other variables) when you access the data. The values of the formula variables do not occupy disk space in the PDB. You can add formula variables to PDB tables by using the IT Service Vision server’s interactive interface or by using the IT Service Vision %CPDDUTL macro.
10. Select Advanced and provide the subject and format, if applicable. In this case, select the right arrow that is associated with the Format field, date7., and OK. Edit the format to DATE. and then select OK. Note that the new formula variable is now on the list of variables. Repeat these steps to define another formula variable, UDAY, which has the formula: UDAY = weekday(datepart(datetime)); and the description number of the day of the week (1-7), where 1=Sunday , and defaults for the remaining values.
Cabletron SPECTRUM Appendix 4: Defining IT Service Vision Tables for SPECTRUM Data There are a number of methods for defining IT Service Vision tables for SPECTRUM data. For instance, IT Service Vision supplies table definitions for many common SPECTRUM model types. You can use those table definitions directly (by adding them from the master data dictionary to the PDB’s data dictionary). Or you can use those table definitions as templates for a table definition for a different model type.
MODTYP=specgway.modtyp where specgway is the library name that you defined, before running this macro, by using the libname statement. specgway should point to the directory to which you exported SPECTRUM data. Note: Each observation in the model type data set maps one model type handle to the corresponding model type name. ATTR= This parameter identifies the attributes data set that you exported using SPECTRUM. Use a two-level name in the form libref.
WHICHAID=attrid in (65134 65980 64900) selects only the variables with the indicated values of attrid. WHICHMTH= This parameter is used to subset the data so that the resulting table definitions reflect only the model types that you select. The value of this parameter is a list of the model types that should be used for creating tables. For example: WHICHMTH=mth in (65134 65132) selects only those device types for table creation.
Cabletron SPECTRUM Appendix 5: Editing Model Types for SPECTRUM 1. Stop SpectroSERVER. To edit models in SPECTRUM, you will need to stop SpectroSERVER. Using the SPECTRUM Control Panel, stop the SpectroSERVER by selecting Stop SERVER. Select OK to confirm that you want to shut the server down. The server is stopped when messages similar to the following appear in the message window of the Control Panel Closing /usr/spectrum/SS/SpectroSERVER database...
a. To change the logging flag for a given attribute, select the attribute name from the list. The attribute selected should be one of the variables in the corresponding IT Service Vision table. For example, select avgBusy1. b. To change the attribute’s flag, select File -> Open Attribute... SPECTRUM displays the Attribute Edit View window. c. To enable logging, depress the Logged radio button in the Extended Flags section of the window. d. To save the change, select: File -> Save & Close Attr Edit.
Cabletron SPECTRUM Appendix 6: Mapping SPECTRUM Data Types to an IT Service Vision PDB To read SPECTRUM data into an IT Service Vision PDB, it is necessary to tell SPECTRUM to write out the same data that you tell the IT Service Vision server to read in. If what is written does not correspond to what is read, the result is generally that %CSPROCES reads no data (or a smaller amount of data than you expected) into your PDB.
NOTE: NOTE: NOTE: NOTE: NOTE: +--------------------------------------------------------------| Table AgeLim ObsIn ObsKept RBDF LastIn First-LastKept | | CCIRTR 10 0 0 N/A N/A N/A +--------------------------------------------------------------- then the model types that you exported do not match the IT Service Vision tables that you named on the %CSPROCES macro (or selected, if you defined the process step through the server’s interactive interface), or the attributes that you exported do not match the var
/* Map SPECTRUM model handle to model type handle */ data _null_; set specgway.model end=lastrec; file temp; if _n_=1 then put ’proc format; value mh2mth’; put mh ’=’ mth; if lastrec then put ’; run;’; run; %include temp; and submit the program (by typing SUBMIT on the command line and pressing ENTER or RETURN) after the %CSPROCES macro completes. a. When the program completes (as indicated in the SAS log), issue this command fsview specgway.stat b. Type bye.
and press ENTER or RETURN. The values for MH now display as model type handles. Compare this number to the ID number of the corresponding IT Service Vision table. To check table ID numbers, follow this path from the IT Service Vision main window: Administration -> Manage Tables -> select a table -> click right mouse button -> Properties and look in the ID Number field, which is approximately in the middle of the window. The table’s ID Number needs to match the model type handle in the specgway.
that are to be exported; for example: TCP_Retrans_Min_Val 1007f These attribute ID numbers should match the attribute ID numbers in the actual exported statistics data set and in your PDB’s data dictionary. Keep this window up so that you can compare attribute ID numbers from these other sources in the next steps. 3.
Variable -> Collector: Selected -> SPECTRUM -> OK -> OK -> scroll to the variable and select it -> click right mouse button -> Copy Variable to Current Table -> Close IT Service Vision re-displays the list of variables in the table, and the new variable is now in the list. Select Close -> Close -> OK to return to the main window. Alternatively, you can add the variable to the table by using %CPDDUTL control statements.
OpenView and NetView Appendix 1: Recommended Metrics Note: The tables whose names start with HN2 in IT Service Vision contain MIB-II data collected by HPOV/NVAIX. This document describes working with tables HN2IFT and HN2NIX. MIB-II Table/Group Name Which Metrics PDB Table Name mib-II.egp All metrics (1.3.6.1.2.1.8.*) HN2EGP mib-II.icmp All metrics (1.3.6.1.2.1.5.*) HN2ICM mib-II.interfaces All metrics (1.3.6.1.2.1.2.1) HN2INF mib-II.ip All metrics HN2IP_ (1.3.6.1.2.1.4.
OpenView and NetView Appendix 2: Alternate Data-Logging Method This is an alternate version of the task in Section 1, Task 3: Create a Test PDB and Process, Reduce, and Report on Your Data In this version of the task, you create your own snmpCol.conf file (the file of variables for which data are to be collected and the nodes on which the data are to be collected). You also manually stop and start the snmpCollect daemon.
4. Create an snmpCol.conf file. Issue this command: /tmp/csgmkcnf 5. Back up the existing /usr/OV/conf/snmpCol.conf file. The existing snmpCol.conf file contains your current collection configuration. (If you made any modifications in the configuration, see Section 1, Task 3: Create a Test PDB and Process, Reduce, and Report on Your Data , the changes that you made are included in this file. In this step you will save those changes but not use them.) Issue this command: cd /usr/OV/conf cp -p snmpCol.
on dimes.unx.sas.com because dimes is not a router. The snmpCollect daemon ignores this metric for dimes and proceeds normally with collecting the rest of the metrics for dimes. Another kind of message you may see is like this: Tue Jun 06 11:41:40 1995 : swimmer.unx.sas.com doesn’t reply to a 100 object but responds to sysUpTime. Be sure SNMP timeouts are not set too small (SNMP interval: 0.80s retry: 3).
from the main menu in OVW: Options -> SNMP Configuration -> Global default -> change the value in the timeout -> Replace -> OK
OpenView and NetView Appendix 3: OVW-Specific Tips for Reporting Subsetting by Using a Format in Local or Global Where If you have too many values of interest to fit in a WHERE expression or if a variable needs some sort of transformation (for example, a table lookup) before you can construct a WHERE expression, you may want to use a user-written format.
To create the $USVCS format: 1. Make the SAS PROGRAM EDITOR window the active window. 2. Issue this command include !sasroot/misc/cpe/csgfmt2 SAS displays the program in the body of the PROGRAM EDITOR window. The program has a comment that describes the PDB table and level on which it runs to create the format. In this case, the program is designed to run on SNMP MIB-II data in the DETAIL.HN2NIX table. 3. To modify the program so that the format is saved in SITELIB.
Vision. Note: The appropriate table to add the formula variable to is the table that contains the variables on which the formula variable is based. In addition to making the Kept status of the formula variable Yes, remember to check that the variables on which the formula variable is based also have Kept status set to Yes. Adding a Formula Variable By Using the %CPDDUTL Macro IT Service Vision supplies a example program for using the %CPDDUTL macro to define formula variables.
Administration -> Manage Tables -> select the table -> click right mouse button -> Properties -> General -> Variables 2. For one formula variable, for instance UWEEK, select Tools and Create Formula. IT Service Vision displays the Create New Formula window. Type in the variable name UWEEK and select OK. IT Service Vision displays the Make/Edit Formula window. Select General. 3. Select the levels at which you want the formula calculated.
SunNet Manager and Enterprise Manager Appendix 1: Mapping SunNet Manager Metrics to IT Service Vision Tables and Variables SunNet Manager logged data is read into IT Service Vision tables based on the name of the table.
SunNet Manager and Enterprise Manager Appendix 2: Defining Tables from Your Own Schema Files To read data logged by SunNet Manager using a schema file not supported by IT Service Vision, you must first create IT Service Vision dictionary entries for the tables of metrics defined in the schema file using the %CSSNMSCH and %CPDDUTL macros. Run these macros iteratively to create suitable definitions in your PDB for the tables in the schema file.
Create a file that maps long names to short names. Use the messages produced in the preceeding step and the table and variable name definitions in !SASROOT/misc/cpe/cssnmnam as a guide (note that the message produced by %CSSNMSCH in the SAS log are of a form that they can be cut and pasted directly into the file that maps long names to short names. Just type over UNKNOWN with a correct name. There are rules regarding table and variable names for IT Service Vision.
4. Define the tables and examine/correct the definitions Despite the best efforts of %CSSNMSCH, some schema files do not convert perfectly to IT Service Vision dictionary definitions. Run %CPDDUTL on the dictionary update statements created above to define the new entries in your PDB (a new PDB presumably created just for this purpose). Examine the definitions using the IT Service Vision interactive interface and compare them to the schema files.
