Patch Name: PHSS_29383 Patch Description: s700_800 11.X OV ITO6.X HP-UX 11.X OV-PA Agent C.03.71 Creation Date: 03/07/17 Post Date: 03/08/01 Hardware Platforms - OS Releases: s700: 11.00 11.11 s800: 11.00 11.11 Products: OpenView IT/Operations 6.0 Filesets: OVOPC-PA-CLT.OVOPC-PA-HP11,fr=A.06.00,fa=HP-UX_B.11.00_32/64,v=HP OVOPC-PA-CLT.OVOPC-PA-HP11,fr=A.06.00,fa=HP-UX_B.11.11_32/64,v=HP Automatic Reboot?: No Status: General Release Critical: No Category Tags: defect_repair general_release manual_dependencies Path Name: /hp-ux_patches/s700_800/11.X/PHSS_29383 Symptoms: PHSS_29383: - 8606281891: HPUX: APP_CPU_TOTAL_UTIL sum too low compared to GBL_CPU_TOTAL_UTIL. - 8606293100: sdlcomp has problems with leading zeros in spec files. - R555012140: 11x: PROC_TTY, THREAD_TTY shown as "255 0xffff" for procs w/o terminal. - R555016793: HPUX: intermittent invalid PROC_CPU_*_UTIL > max possible. - R555021280: MP: FS data summarization (extract -xp -Y) does not work. PHSS_28380: - 8606221610: Metric help for the metric GBL_NET_OUTQUEUE says that this metric is computed from the sum of the BYNETIF_QUEUE metrics for the various LANs that are on the system. - 8606236261: With a blank authip file established in /var/opt/perf, connection to OVPA from a client on the same system is refused with an "UNAUTHORIZED CONNECTION ATTEMPT" entry in the status.rep_server or status.alarmgen file. This problem only occurs on multi-homed or ServiceGuard cluster systems. - 8606248354: Blank process names are displayed by glanceplus and logged by the performance agent. The amount of transaction data logged decreases (independent of load). And, messages of the form: WARNING: Transaction Overflow. (PE221-49) are repeatedly written to status.scope. - 8606256903: The utility -xr command can produce resized logfiles that can be significantly smaller than expected if the days= option is used. - D500344101: Enhance MeasureWare so that PROC_MEM_VIRT (virtual memory size) for a process can be used as a threshold filter in the parm file in order to consider a process "interesting" from the standpoint of scopeux process logging. - H554002240: MWA (VPPA) utility RESIZE command can empty LOGGLOB when: - resizing LOGAPPL to a very small value and then resizing LOGGLOB results in an empty LOGGLOB (with PARM file setting SIZE=0.0) - resizing LOGGLOB to a very small value and then resizing LOGAPPL results in an empty LOGGLOB (with PARM file setting SIZE=0.0) - H554002351: The utility -xr command can produce resized logfiles that can be significantly smaller than expected if the days= option is used. - R555010654: On HPUX 11.x and beyond, the performance tools are intermittently missing process death records. This also can show up in the Glanceplus Process List screen when a dead process continues to show up as an active process. - R555016345: OVPA is under reporting the disk activity in the BYDSK_PHYS_*_BYTE_* metrics during a 5 minute interval. - R555018735: The FILESYSTEM class discriminater metric has been changed from FS_DEVNAME (device special file name) to FS_DIRNAME (mount point path name). This change will cause PerfView to show filesystem data selection per mount point, and will affect the ability to export pre-existing FILESYSTEM class summary records (see R555017237 in "Known Problems and Workarounds", Chapter 4). PHSS_27204: - R555013479: OV Performance Agent would sometimes present incorrect values for the process metrics PROC_IO_BYTE_CUM, PROC_DISK_PHYS_IO_CUM and PROC_DISK_LOGL_IO_CUM. - R555013633: OV Performance Agent products now support IPv6 network interfaces. On HP-UX 11.11, with IPv4/IPv6 (dual-stack) installed and configured, the following error message should no longer occur on a regular basis in status.scope: Can't get ID_ifEntry /dev/ip6 ppa_num = 0 Invalid argument. - R555013938: Source modifications have been made that significantly reduce the likelihood of an midaemon core dump when out-of-order KI traces are processed. Out-of-order KI traces are due to unsynchronized CPU clocks that may occur on some multi-CPU "L" and "N" class systems. This midaemon enhancement simply reduces the risk of midaemon cores on systems with suspect hardware. Reference SR# R555013951 in the "Known Problems and Workarounds" section for more information on the underlying cause. - R555014009: Previously, bad network metric values had been observed on busy Omniback servers, and the bad values were possible on any system with very high network throughput rates on the order of thousands of packets per second. - R555014699: A problem was fixed that could cause intermittent scope cores during execution of a "mwa stop" command. - 8606228123: A problem was corrected where BYNETIF_COLLISION could show zero for all interfaces, yet GBL_NET_COLLISION_PCT could be greater than zero. - 8606230092: The mettable.txt file was confusing because it was not clear to which versions the record layouts applied. To fix this problem, OV Performance Agent version strings have been added to record descriptions for all versions C.00.00 and newer. PHSS_26323: - R555005915 MP: scope derives PROC_CPU_TOTAL_TIME_CUM incorrectly - R555011258 HP 11.0 perflbd may hang/abort preventing connection without OS patch - R555011267 Idle diaglogd process shows up blocked on "IO" in HP-UX 11.11: kslp fix - R555011405 MP:GBL_DISK_PHYS_BYTE_RATE and related metrics overflow in 5min interval - R555011701 perflbd failure on HP-UX 11.11 when in IPv4 - IPv6 "dual stack" mode - R555012572 HP: Blocked state mapping for "async" I/O bound processes is VM s/b DISK - R555012617 HP: 11.x GBL_NET_OUTQUEUE reporting incorrect values - R555013222 HP: 11.0 BYNETIF_ID for lan0 and lan1 is 1 on system; needs libnm patch - R555013314 MP: intermittent scopeux failure and core dump - R555013625 exported binary filesystem and transaction data both use record id 12 - R555013647 HP: 11.11 midaemon cores with VxVM 3.2 configured - 8606220049 HP: 11.11 performance tools network metrics not reported with PHNE_24211 - 8606224561 Glance - GBL_NFS_CLIENT_WRITE_BYTE_RATE is 0 when using NFS version 3 PHSS_25668: - 8606163706 Various process memory metrics and process memory region metrics do not report values greater than 4 GB. - 8606165364 Data for classes with multiple instances (i.e. all except GLOBAL) is summarized incorrectly by the export function of extract when a shift is specified and the end of the shift occurs before the end of the summarization interval. - 8606200749 In Glance, the "Network By Interface" metric screen values should consist of only physical numeric metrics values for HPUX 11.0. (JAGad69925) - R555010638 PROC_CPU_TOTAL_UTIL_CUM can generate invalid values. This can be seen when monitoring a multi-threaded cpu intensive process. The value will drop due to an overflow. - R555010312 On HP-UX 11.XX, begining with the C.03.25.00 release of MeasureWare, a known problem exists. During shutdown, perflbd and/or agdbserver may fail to shutdown properly. - R555010653 On systems with a high process creation rate (fork rate), GBL_ALIVE_PROC and PROC_INTERVAL_ALIVE are intermittently incorrect. The values appear to be excessively high. PHSS_24622: - 8606172810 In cases where the PerfView system is no longer available (ip changed, pv removed, etc.) - 8606179953 A process not associated with MeasureWare is terminated by the MeasureWare shutdown script (ie. "mwa stop"). - 8606188240 For systems with >32 disks, extracted logfiles that are created including Global Summary data in the extract using VPPA C.03.20 result in an unreadable extracted logfile. - 8606189304 A memory leak has been detected in C.03.20.00. This leak can be seen when running glance or gpm at a shorter interval for several days. - 8606193105 On B.11.11 the performance tools are accessing NFS direct mounted directories causing the autofs timer to reset. - 8606194442 For glance and gpm version C.03.20 running on HPUX 11.0 the 1, 5 & 15 minute processor load average metrics are presenting only whole number values. - R555005836 Specifying application=all in the parm file produces an application record. - R555007839 Systems with large disk configurations are experiencing unsatifactory overhead when running the performance tools. - R555008630 The prompts output from the "perfstat -z" command for creating a tar archive of MeasureWare log files are confusing. - R555009961 On HPUX 11.x systems the GBL_CPU metrics do not always represent the normalized sum of the BYCPU_CPU metrics. - R555010346 Non-HPUX platform versions of extract attempting to interpret HPUX MeasureWare version C.03.20 /C.03.25 application data may experience some unusual values for PRM specific metrics. - R555010390 On hpux 11.11 releases of the performance tools the PRM_CPU metrics are incorrect. - R555010424 It is possible that new process and killed process events will fail to match. These are, respectively, Interest Reasons 'N' and 'K'. PHSS_23972: - 8606168043 The metric BYDSK_AVG_SERVICE_TIME is not displayed as a real number with two decimal places of precision. - 8606178177 Whitespace characters trailing application names in the parm file can result in 'visually duplicate' application names in logidx for OVPA. These 'duplicates' also appear in Perfview, extract, etc. - R555009639 The MPE metrics GBL_CPU_TOTAL_UTIL and GBL_CPU_TOTAL_TIME are reporting "missing" when an export of MPE extracted logs is performed with MPE test file "rxday" which contains 1 CPU worth of data. - R555009970 In Measureware versions C.02.60 and greater have changed the ARM "application response and measurement" libraries to use pthreads rather than user/dce threads. - R555010113 On HP-UX 11.0 the performance tools can experience intermittent SIGBUS failures. - R555010145 HP-UX MeasureWare cannot read application data from Sun MeasureWare logfiles from Sun MeasureWare versions C.02.53 or C.03.10. - R555010185 HP-UX MeasureWare version C.03.20 cannot access process data from Sun logs written with version C.03.10 - R555010195 HP-UX MeasureWare version C.03.20 cannot access tansaction data from Sun logs written with version C.03.10 or AIX logs written with version C.03.15. PHSS_22483: - 5003449975 MeasureWare extract appends duplicate records to extracted log. - 5003456681 LV_SPACE_UTIL not updated when file system is expanded. - 8606107840 Cannot connect to MeasureWare through a NAT firewall. - 8606157484 Dependencies on transitional links should have been removed, some were missed. - D500397968 Process Resource Manager metrics not supported by MeasureWare. - R555007835 Incorrect extra LAN entries reported with Auto- Port Aggregation. - R555008431 Process cpu metrics incorrect for multi threaded process run with nice. - R555008831 MeasureWare rep_server can get memory exception reading extracted log. - R555009120 MeasureWare extract can get memory exception processing logs with too many devices. - R555009152 MeasureWare extract creates duplicate records extracting NETIF data. Defect Description: PHSS_29383: Resolution: Update version of OVPA deployable subagent to C.03.71. Deployable to HP-UX 11.00 and 11.11 OVO Managed Nodes. - 8606281891: On HPUX, a problem was found where the sum of all instances of the APP_CPU_TOTAL_UTIL metric could be significantly lower than the value of the GBL_CPU_TOTAL_UTIL metric. The problem was reported to occur when the process creation and death rates were high during an interval. However, workloads still exist where the midaemon may need to be configured to start with a larger number of process and thread entries. See the midaemon man page for additional information. - 8606293100: A DSI problem was discovered in the most recent releases of OVPA, where metric numbers containing leading zeros caused the sdlcomp compiler to fail with an error message indicating that "Metric must be between 1 and 999999." - R555012140: On HP-UX, Glance previously returned 255 0xffff for PROC_TTY and THREAD_TTY for processes/threads that have no terminal. These metrics now return blank when there is no terminal associated with the process/thread. The changes for PROC_TTY are available for OVPA as well. - R555016793: On HP-UX, the PROC_* metrics class has been corrected to properly account for dying threads in a multithreaded process. This will resolve the problem where the PROC_CPU_*_UTIL metrics exceeded the maximum possible. - R555021280: Fixed a problem that made it impossible to extract/export filesystem summary data or graph filesystem data with OV Performance Manager. PHSS_28380: Resolution: Update version of OVPA deployable subagent to C.03.70. Deployable to HP-UX 11.00 and 11.11 OVO Managed Nodes. PHSS_27204: check the list of the symptoms Resolution: Update version of OVPA deployable subagent to C.03.58. Deployable to HP-UX 11.00 and 11.11 OVO Managed Nodes. PHSS_26323: check the list of the symptoms Resolution: Update version of OVPA deployable subagent to C.03.55. Deployable to HP-UX 11.00 and 11.11 OVO Managed Nodes. PHSS_25668: check the list of the symptoms Resolution: Update version of OVPA deployable subagent to C.03.50. Deployable to HP-UX 11.00 and 11.11 OVO Managed Nodes. PHSS_24622: check the list of the symptoms Resolution: Update version of OVPA deployable subagent to C.03.35. Deployable to HP-UX 11.x OVO Managed Nodes. PHSS_23972: check the list of the symptoms Resolution: Update version of OVPA deployable subagent to C.03.25. Deployable to HP-UX 11.00 and 11.11 VPO Managed Nodes. PHSS_22483: check the list of the symptoms Resolution: Update version of OVPA deployable subagent to C.03.20. Deployable to HP-UX 11.00 and 11.11 VPO Managed Nodes. Enhancement: No SR: R555021280 8606293100 8606281891 R555016793 R555012140 R555016345 R555018735 R555010654 8606236261 H554002240 8606248354 8606256903 H554002351 8606221610 D500344101 8606228123 R555013633 R555013479 R555014009 8606230092 R555014699 R555013938 8606224561 R555005915 R555011405 8606220049 R555013625 R555012572 R555013222 R555011701 R555013314 R555011258 R555013647 R555011267 R555010653 8606165364 R555010638 8606200749 8606163706 R555010312 R555007839 R555010424 R555008630 R555005836 8606179953 8606193105 R555010346 R555009961 8606194442 8606172810 R555010390 8606188240 8606189304 8606178177 R555010145 R555009639 R555010113 R555009970 R555010185 R555010195 8606168043 R555009152 R555008431 D500397968 8606107840 5003456681 5003449975 R555009120 8606157484 R555007835 R555008831 Patch Files: OVOPC-PA-CLT.OVOPC-PA-HP11,fr=A.06.00, fa=HP-UX_B.11.00_32/64,v=HP: OVOPC-PA-CLT.OVOPC-PA-HP11,fr=A.06.00, fa=HP-UX_B.11.11_32/64,v=HP: /var/opt/OV/share/databases/subagent/VP_Perf_Agt/hp/pa-risc/ hp-ux11/SubAgentPlatform /var/opt/OV/share/databases/subagent/VP_Perf_Agt/hp/pa-risc/ hp-ux11/C.03.71/install/ovpa_inst /var/opt/OV/share/databases/subagent/VP_Perf_Agt/hp/pa-risc/ hp-ux11/C.03.71/ovpa_pkg.Z /var/opt/OV/share/databases/subagent/VP_Perf_Agt/hp/pa-risc/ hp-ux11/C.03.71/ovpa_pkg.Z.B.11.11 what(1) Output: OVOPC-PA-CLT.OVOPC-PA-HP11,fr=A.06.00, fa=HP-UX_B.11.00_32/64,v=HP: /var/opt/OV/share/databases/subagent/VP_Perf_Agt/hp/pa-risc/ hp-ux11/SubAgentPlatform: None /var/opt/OV/share/databases/subagent/VP_Perf_Agt/hp/pa-risc/ hp-ux11/C.03.71/install/ovpa_inst: rsubagtinst C.03.71.00 10/06/03 =*= /var/opt/OV/share/databases/subagent/VP_Perf_Agt/hp/pa-risc/ hp-ux11/C.03.71/ovpa_pkg.Z: $Revision HP-UX 11.00 04Jan02 dhynie$ $Revision HP-UX 11.00 19Mar01 danielh $ $Revision HP-UX 11.00: 20MAY00 dhynie $ $Revision HP-UX 11.00: 17May00 dhynie $ $Revision HP-UX 11.00: 16Jan01 dhynie $ $Revision HP-UX 11.00: 17May00 dhynie $ $Revision HP-UX 11.00: 18Dec00 dhynie $ $Revision HP-UX 11.00: 19Mar01 danielh $ $Revision HP-UX: 10.20 & 11.00 01Jun00 dhynie $ $Revision HP-UX: 10.20 & 11.00 01Jun00 dhynie $ $Revision HP-UX: 11.00 06Sep02 dhynie $ $Revision HP-UX 11.x 22Mar02 dhynie $ $Revision HP-UX: 11.00 06Sep02 dhynie $ $Revision HP-UX: 10.20 & 11.00 01Jun00 dhynie $ $Revision HP-UX: 10.20 & 11.00 16Sep02 dhynie $ $Revision HP-UX 11.00: 06Sep02 dhynie $ $Revision HP-UX: 1100 17Jul02 dhynie $ $Revision HP-UX 10.20 & 11.00: 01Jun00 dhynie $ $Revision HP-UX 10.20 & 11.00: 01Jun00 dhynie $ $Revision HP-UX: 10.20 & 11.00 01Jun00 dhynie $ /var/opt/OV/share/databases/subagent/VP_Perf_Agt/hp/pa-risc/ hp-ux11/C.03.71/ovpa_pkg.Z.B.11.11: $Revision HP-UX 11i: 21Sep00 dhynie $ $Revision HP-UX 11i 04Jan02 dhynie$ $Revision HP-UX 11i: 19Mar01 danielh $ $Revision HP-UX 11i: 20MAY00 dougg $ $Revision HP-UX 11i: 15May00 dhynie $ $Revision HP-UX 11i: 13Mar03 dhynie $ $Revision HP-UX 11i-1122: 08Oct02 dhynie $ $Revision HP-UX 11i-1122: 24Apr02 dhynie $ $Revision HP-UX 11i: 19Mar01 danielh $ $Revision HP-UX 11i: 19Mar01 danielh $ $Revision HP-UX 11i: 15May00 dhynie $ $Revision HP-UX 11i: 06Sep02 dhynie $ $Revision HP-UX 11.x 22Mar02 dhynie $ $Revision HP-UX 11i: 06Sep02 dhynie $ $Revision HP-UX 11i: 15May00 dhynie $ $Revision HP-UX: 11i 06Sep02 dhynie $ $Revision HP-UX: 11.11 17Jul00 dhynie $ $Revision HP-UX: 11i 17Jul02 dhynie $ $Revision HP-UX 11i: 15May00 dhynie $ $Revision HP-UX 11i: 15May00 dhynie $ $Revision HP-UX 11i: 15May00 dhynie $ cksum(1) Output: OVOPC-PA-CLT.OVOPC-PA-HP11,fr=A.06.00, fa=HP-UX_B.11.00_32/64,v=HP: 1923402190 745 /var/opt/OV/share/databases/subagent/ VP_Perf_Agt/hp/pa-risc/hp-ux11/SubAgentPlatform 3559942756 5142 /var/opt/OV/share/databases/subagent/ VP_Perf_Agt/hp/pa-risc/hp-ux11/C.03.71/install/ ovpa_inst 247872050 18749440 /var/opt/OV/share/databases/subagent/ VP_Perf_Agt/hp/pa-risc/hp-ux11/C.03.71/ovpa_pkg.Z 965959983 16875520 /var/opt/OV/share/databases/subagent/ VP_Perf_Agt/hp/pa-risc/hp-ux11/C.03.71/ ovpa_pkg.Z.B.11.11 Patch Conflicts: None Patch Dependencies: None Hardware Dependencies: None Other Dependencies: In order to be able to deploy the OVPA subagent to a HP-UX 11.11 system, you need to install the OVO server patch A.06.03 (PHSS_22761 on HP-UX 10.20, PHSS_22762 on HP-UX 11 or ITOSOL_00061 on Solaris) or a superseding patch. This patch is not necessary, if you deploy the OVPA subagent only to systems with HP-UX 11.00. Supersedes: PHSS_28380 PHSS_27204 PHSS_26323 PHSS_25668 PHSS_24622 PHSS_23972 PHSS_22483 Equivalent Patches: ITOSOL_00243: sparcSOL: 2.6 2.7 2.8 Patch Package Size: 33890 KBytes Installation Instructions: Please review all instructions and the Hewlett-Packard SupportLine User Guide or your Hewlett-Packard support terms and conditions for precautions, scope of license, restrictions, and, limitation of liability and warranties, before installing this patch. ------------------------------------------------------------ 1. Back up your system before installing a patch. 2. Login as root. 3. Copy the patch to the /tmp directory. 4. Move to the /tmp directory and unshar the patch: cd /tmp sh PHSS_29383 5. Run swinstall to install the patch: swinstall -x autoreboot=true -x patch_match_target=true \ -s /tmp/PHSS_29383.depot By default swinstall will archive the original software in /var/adm/sw/save/PHSS_29383. If you do not wish to retain a copy of the original software, include the patch_save_files option in the swinstall command above: -x patch_save_files=false WARNING: If patch_save_files is false when a patch is installed, the patch cannot be deinstalled. Please be careful when using this feature. For future reference, the contents of the PHSS_29383.text file is available in the product readme: swlist -l product -a readme -d @ /tmp/PHSS_29383.depot To put this patch on a magnetic tape and install from the tape drive, use the command: dd if=/tmp/PHSS_29383.depot of=/dev/rmt/0m bs=2k Special Installation Instructions: PHSS_27204: This patch changes how Glance is deployed from the OVO Management Server. Previously, a new trial version of Glance was always deployed, whether Glance previously existed on the system or not. Now, Glance will only be deployed and updated if a previous version was already on the system. If Glance was not installed before, then it will not be deployed.