ONCplus B.11.31.
© Copyright 2009, 2012 Hewlett-Packard Development Company, L.P Legal Notices Confidential computer software. Valid license required from Hewlett-Packard for possession, use or copying. Consistent with FAR 12.211 and 12.212, Commercial Computer Software, Computer Software Documentation, and Technical Data for Commercial Items are licensed to the U.S. Government under vendor’s standard commercial license. The information contained herein is subject to change without notice.
Contents 1 ONCplus Overview....................................................................................4 Overview................................................................................................................................4 Availability of ONCplus as an ISU..............................................................................................4 NIS+ Obsolescence Notice........................................................................................................
1 ONCplus Overview This chapter addresses the following topics: • “Overview” (page 4) • “Availability of ONCplus as an ISU” (page 4) • “NIS+ Obsolescence Notice” (page 5) • “NIS Protocol Version 1 Obsolescence Notice” (page 5) • “Related Information” (page 5) Overview Open Network Computing (ONC) comprises core services that enable administrators to implement distributed applications in a heterogeneous distributed computing environment. It also includes tools to administer clients and servers.
NIS+ Obsolescence Notice Network Information Service Plus (NIS+) is a distributed database system that enables the maintenance of commonly used configuration information on a master server and propagates the information to all the hosts in the network. Starting with HP-UX 11i v3, NIS+ is no longer supported. Hewlett-Packard recommends that users migrate to LDAP. NOTE: For information on how to migrate from NIS+ to LDAP, see the NIS+ to LDAP Migration Guide.
2 ONCplus B.11.31.15 This chapter contains the most recent product information pertaining to the Open Networking Computing (ONC) product, version B.11.31.15, which is supported on the HP-UX 11i v3 operating system. This chapter addresses the following topics: • “Announcement” (page 6) • “What's New in ONCplus B.11.31.15” (page 6) • “Enhancements and Fixes in ONCplus to date” (page 6) • “Installation Information ” (page 13) Announcement This version of ONCplus (B.11.31.
Table 1 ONCplus Enhancements and Software Defect Fixes (continued) CR ID Description ONCplus Version QXCR1001162950 11.31 KLM - a cancel/grant race issue. B.11.31.15 QXCR1001162529 11.31 KLM4 lm_block_lock has memory leak. B.11.31.15 QXCR1001189599 Auotofs direct map entry misleads by missing mount options in mnttab. B.11.31.15 QXCR1001190991 11.31 system panic in hpnfs3_rdwr/nfs3_async_dio_read. B.11.31.15 11.
Table 1 ONCplus Enhancements and Software Defect Fixes (continued) CR ID Description QXCR1001127336 The locking mechanism for the e_cnt member of the endpnt_type function has been modified to restrict the maximum number of endpoints. B.11.31.13 QXCR1001131929 The ‘mount -o vers=4,proto=tcp...’ command may fail when a firewall is present. A high priority thread may get into a loop by waiting for a lock that is held QXCR1001044409 by a lower priority thread, which may result in a non-pingable hang.
Table 1 ONCplus Enhancements and Software Defect Fixes (continued) CR ID Description ONCplus Version The share_nfs(1M) manpage does not explain that NFS-serv shared filesystems QXCR1001028325 cannot be unmounted. B.11.31.10 QXCR1001004707 HP-UX 11i v3 NFS pv4 readdir request size is larger than nfs4_bsize setting. B.11.31.10 The rpc.mountd daemon denies mount requests due to incorrect netmask QXCR1000985674 matching. B.11.31.09.
Table 1 ONCplus Enhancements and Software Defect Fixes (continued) CR ID Description ONCplus Version QXCR1000982019 HP-UX 11i v3 rpc.pcnfsd fd leaks. B.11.31.09 The rpc.pcnfsd daemon manipulates effective uid in thread in an unsafe QXCR1000982421 manner. B.11.31.09 QXCR1000982592 The rpc.pcnfsd daemon exits when it is heavily loaded. B.11.31.09 QXCR1000987966 Possible memory corruption due to buffer overflow. B.11.31.09 Incorrect usage of the 'pthread_getspecific' API leads to the rpc.
Table 1 ONCplus Enhancements and Software Defect Fixes (continued) CR ID Description ONCplus Version - Data corruption during READ, when the nfs3_bsize tunable value is greater than 64 KB QXCR1000797858 NFSv3 and NFSv4 set the transfer size incorrectly. B.11.31.06 Compatibility issue with HP-UX 11i v2 xdr_callmsg(), xdr_callhdr(), QXCR1000828306 and xdr_replymsg(). B.11.31.06 QXCR1000864699 NFS client data corruption due to out-of-order write replies. B.11.31.
Table 1 ONCplus Enhancements and Software Defect Fixes (continued) CR ID Description ONCplus Version QXCR1000764069 Added support for SecureNFS in a Serviceguard environment. B.11.31.03 QXCR1000764257 An NFS server panics when using a public filehandle for a lookup. B.11.31.03 QXCR1000771618 B.11.31.03 There is a potential race condition on an NFSv4 client using direct I/O that can result in a hang. QXCR1000770472 NFS READDIRPLUS returns no entries when rsize is set to a very low value. B.11.31.
Table 1 ONCplus Enhancements and Software Defect Fixes (continued) CR ID Description ONCplus Version QXCR1000585874 The nfsstat command does not display access, close, and commit values for NFSv4 when used with the -c or -s options . B.11.31.01 QXCR1000585903 Using kctune to change NFSv4 parameters causes a panic. B.11.31.01 QXCR1000585959 Users cannot access an NFS mounted file system in PAM enabled Kerberos systems. B.11.31.01 QXCR1000586940 The xdr_setpos function in libnsl.
ONCplus ONCplus ONCplus ONCplus ONCplus ONCplus ONCplus 3. B.11.31.06 B.11.31.05 B.11.31.04 B.11.31.03 B.11.31.02 B.11.31.01 B.11.31 ONC+ ONC+ ONC+ ONC+ ONC+ ONC+ ONC+ 2.3 2.3 2.3 2.3 2.3 2.3 2.3 After you download the ONCplus_B.11.31.15.depot file, move it to the /tmp directory: /tmp/ONCplus_B.11.31.15.depot 4. Verify that the file has downloaded correctly using the swlist command as follows: swlist -d @ /tmp/ONCplus_B.11.31.15.
4. To revert to the previous version of ONCplus enter the following command: swinstall -x allow_downdate=true -x autoreboot=true \ -s /tmp/ONCplus_B.11.31.depot ONCplus NOTE: ONCplus contains kernel filesets. Thus swinstalling the product results in a system reboot after the installation is complete.
WARNING! 1. Problem There is a possibility for an NFS client to experience a system hang if you revert to one of the below ONCplus versions: • ONCplus B.11.31.09 • ONCplus B.11.31.09.01 • ONCplus B.11.31.09.02 Cause The NFS client could hang under below conditions/configurations: • Severe file cache memory pressure due to low file cache configuration. For example, filecache_min = filecache_max =1% • When the nfs3_max_threads tunable is set to “0”.
3. 4. 5. On an HP-UX server with ONCplus B.11.31.07.01 installed, NFSv3 does not fully support large files. Directory related operations on an NFS client with ONCplus B.11.31.06 or B.11.31.07 versions installed along with mounted file systems with read/write permissions that are greater than 8192 bytes in size may result in system panic or data corruption. If your system has ONCplus B.11.31.06 or B.11.31.07 version installed, HP recommends that you install ONCplus B.11.31.07.01 version.
7. ONCplus B.11.31.02 may cause a boot panic situation if you are installing on an HP 9000 rp3410 or rp3440 system with the September 2007 release of HP-UX already installed. If your system has this configuration, Hewlett-Packard recommends that you install the March 2008 release of HP-UX, which includes ONCplus B.11.31.02, or install ONCplus B.11.31.06. You can use the "model" command to print the hardware model information.
8. Problem • The write operation may be interrupted when very large files are copied over NFS. • System crash may occur under heavy load conditions. Solution This problem has been fixed in ONCplus B.11.31.14. Workaround For ONCplus B.11.31.13, the workaround is to use the forcedirectio option for the NFS mount in case of loopback mount or cross mounts with very less UFC cache (say 1%) configured on NFS client systems.
On ONCplus B.11.31.14 version, Fix was implemented only for NFSv3 client version only. To enable this fix on this version, following tunables must be set to the below values: kctune nfs_enable_write_behind = 1 • kctune nfs_enable_ufc_threshold = 1 • kctune nfs3_ufc_threshold_percentage = 50 NOTE: All NFS v3 filesystems have to be unmounted if there is any change in these tunables or UFC filecache_max. Workaround For ONCplus B.11.31.
NOTE: Doing a force unmount may cause data loss for open files. After the unmount is complete, if a program tries to access files from an unmounted filesystem, it will get an EIO error. For information on HP-UX SRP commands and HP-UX Containers, see the HP-UX Containers (SRP) Administrator's Guide. To locate this document, go to the following location on the HP Business Support Center: www.hp.com/go/virtualization-manuals. On this page, select HP-UX Containers (SRP) Software.
3 Features Introduced in previous versions of ONCplus This chapter describes the features introduced and/or defect fixes in all the previous versions of ONCplus B.11.31. • “Features Introduced in ONCplus B.11.31.14” (page 22) • “Features Introduced in ONCplus B.11.31.13” (page 22) • “Features Introduced in ONCplus B.11.31.12” (page 22) • “Features Introduced in ONCplus B.11.31.11” (page 23) • “Features Introduced in ONCplus B.11.31.10” (page 23) • “Features Introduced in ONCplus B.11.31.09.
When ONCplus is operating in the SRP environment, the following features/services are supported: • NFSv2, NFSv3 and NFSv4 client services. • AutoFS service. • RPC port mapper service daemon (rpcbind): All the RPC user space daemons and libraries can operate in the SRP environment. • Network Lock manager (NLM) and Network Status Monitor (rpc.lockd and rpc.statd) client-side services. • Generating NFS client mount point statistics using the nfsstat command. • NIS server and client services.
Features Introduced in ONCplus B.11.31.09.01 ONCplus B.11.31.09.01 is a defect fix release and does not include any new features. All features introduced in previous ONCplus versions are supported in ONCplus B.11.31.09.01. For information on the defects fixed, see “Enhancements and Fixes in ONCplus to date” (page 6). Features Introduced in ONCplus B.11.31.09 ONCplus B.11.31.09 includes both defect fixes and new features. All features introduced in previous ONCplus versions are supported in ONCplus B.11.31.
a synchronous wait for individual request. This helps in improving the performance of the database with large sequential I/O. This feature can be enabled using the following three tunables. ◦ nfs3_enable_async_directio_read ◦ nfs3_enable_async_directio_write ◦ nfs3_max_async_directio_requests For more information on these tunables, see the Managing NFS and KRPC Kernel Configurations in HP-UX 11i v3 whitepaper. To locate this document, go to the HP-UX Networking docs page at: http://www.hp.
Features Introduced in ONCplus B.11.31.04 ONCplus B.11.31.04 includes both defect fixes and new features. All features introduced in previous ONCplus versions are supported in ONCplus B.11.31.04. The following features were introduced in ONCplus B.11.31.04: New NIS Features in ONCplus B.11.31.04 NIS introduces the following features: • IPv6 support for RPC commands and daemons Phase I ONCplus B.11.31.04 delivers IPv6 support for rpc.rexd, rpc.rwalld, rpc.rusersd, rusers, and rwall commands and daemons.
New NFS Features in ONCplus B.11.31.03 NFS introduces the following features for NFSv4: • File Delegation with Local Access NFSv4 clients support delegation on HP-UX 11i v3. However, until ONCplus B.11.31.03, NFSv4 servers supported delegation with the caveat that no local file access will occur on any delegated file. For example, if the server grants a delegation to the file “/a/b/foo” then any local users on the server need to avoid accessing file “foo” while the delegation is in effect.
by the cache directory (also known as the working set size). This command uses the logfile created by the cachefslog command to display the information. The cachefswssize command, used with the -a option, displays the information in ASCII format. New NIS Features in ONCplus B.11.31.02 NIS introduces the following features: • Support for IPv6 Protocol The NIS clients and servers are now IPv6 enabled.
the server. When another client requests access to the same file, the server revokes the delegation from the first client and grants it to the second client. NOTE: Delegations are disabled by default. If delegations are enabled, they are only supported by applications that access the delegated files remotely via NFS clients. Allowing local access and enabling delegation on a file can corrupt the file.
NOTE: Client failover support is limited to read-only mounts or static filesystems that are not modified often. • Enhanced NFS logging NFS server logging enables an NFS server to provide a record of file operations that are performed on its filesystems. New AutoFS Features in ONCplus B.11.31_LR AutoFS supports the following features: • On-Demand Mounting of Hierarchical Filesystems In earlier versions, AutoFS mounted an entire set of filesystems if they were hierarchically related.
• ◦ HFS ◦ VxFS ◦ CIFS ◦ AutoFS Secure NFS Support If the NFS client supports mounting of secure directories, AutoFS supports Secure NFS filesystems. • IPv6 Support AutoFS provides support for mounting filesystems over IPv6 transports. New CacheFS Features in ONCplus B.11.31_LR CacheFS supports the following features: • Complete Binary Caching CacheFS is commonly used to manage application binaries. This feature enables you to cache a complete binary file in the local cache.
• Switching Mount Options You can switch between mount options without deleting or rebuilding the cache. For instance, you can switch from default to non-shared, or from noconst to demandconst mount options without recreating the cache. • Support for Large Files and Large Filesystems CacheFS supports the maximum file and filesystem sizes supported by both the underlying front filesystem in which the cache resides and the back filesystem. CacheFS data structures are 64-bit compliant.