Patch Name: PHSS_20684 Patch Description: s700_800 11.00 Event Monitoring Services(EMS) patch Creation Date: 00/01/20 Post Date: 00/02/22 Hardware Platforms - OS Releases: s700: 11.00 s800: 11.00 Products: B7609BA EMS A.03.00 B7609BA EMS A.03.00.01 B7609BJ EMS A.03.00 Filesets: EMS-Core.EMS-CORE,fr=A.03.00,fa=HP-UX_B.11.00_32/64,v=HP EMS-Core.EMS-CORE,fr=A.03.00.01,fa=HP-UX_B.11.00_32/64,v=HP EMS-JCore.EMS-CORE,fr=A.03.00,fa=HP-UX_B.11.00_32/64,v=HP EMS-JCore.EMS-CORE,fr=A.03.00.01,fa=HP-UX_B.11.00_32/64,v=HP EMS-Config.EMS-GUI,fr=A.03.00,fa=HP-UX_B.11.00_32/64,v=HP EMS-Config.EMS-GUI,fr=A.03.00.01,fa=HP-UX_B.11.00_32/64,v=HP EMS-JConfig.EMS-GUI,fr=A.03.00,fa=HP-UX_B.11.00_32/64,v=HP EMS-JConfig.EMS-GUI,fr=A.03.00.01,fa=HP-UX_B.11.00_32/64,v=HP Automatic Reboot?: No Status: General Release Critical: No Category Tags: defect_repair general_release Path Name: /hp-ux_patches/s700_800/11.X/PHSS_20684 Symptoms: PHSS_20684: 1. p_client might open an increasing number of files when it restarts itself every 32 minutes (typically). This could also cause an increasing number of registrar processes to be left running. Thus, after a very long time, the system might not be able to open files or run processes. Also, on HP-UX 11.00 only, the system might not be able to automatically assign any more "dynamic" UDP port numbers, if NIS were used as the resolver for gethostbyname(3). PHSS_18435: Same as patch PHSS_18436. PHSS_18436: 1. The EMS GUI running in character mode on 11.0 would not show resources when attempting an "Add" operation. When running on 10.20 or 11.0 in GUI mode, resources would sometimes be missing for the "Add" operation as well . 2. Persistence file grows when client issues rm_query_resource. 3. send_opc_notifcation sends the wrong message when the condition is normal. "Normal" in this case is when the threshold condition is not true and either the initial or return notification is being sent. But in this case, "Problem" is the text for the event parameter instead of normal. 4. startmon_client and psmctd report repeated time-outs in client.log for monitor and subclass requests. 5. Log entries appear to be out of order occassionally, based on timestamp. 6. Log file improvements to help troubleshooting: a. Timestamp has sub-second resolution. b. If /etc/opt/resmon/unlimited_log file is present, disable log file size check, roll-over to log.old, and loss of previous log.old data. 7. "Error renaming p.1.***" error messages from p_client in client.log file. 8. Event 711 error message "Unknown Resource Type" has been logged many times in api.log file. 9. "Bad file number" messages appear in registrar.log, possibly leading to anomalous behavior in clients and miscellaneous errors in client.log. Defect Description: PHSS_20684: 1. p_client might open an increasing number of files when it restarts itself every 32 minutes (typically) because the close-on-exec flag was not set for some file descriptors, and these file descriptors were left open under some error conditions, notably the persistence- lock file and the registrar connection socket. Also, the 11.00 NIS resolver library does not set close-on- exec on its UDP socket descriptor, which is left open in order to optimize subsequent calls. PHSS_18435: Same as patch PHSS_18436. PHSS_18436: Resolution: 1. A pointer value was not being handled correctly which led to the problems 2. The Info Block of a query request should not be written to the persistence file. 3. The test of RmNotificationTrigger is done too late in send_opc_notification(). This causes the test to fail when notification_trigger == RM_NORMAL. 4. registrar gets stuck sleeping in wait() after exec'g p_client, if it had previously exec'g monitors that are still active. 5. Log semaphore is acquired after calling ctime() to create timestamp. 6. Log file improvements: a. Make it easier to see how long activities take when debug logging is enabled. b. Ensure that log data is not lost, particularly useful when debug logging is enabled. 7. p_client was allocating one-less byte than needed for the persistence file path, causing clobbering of data. 8. checking a duplicated monitor request does not work correctly when the operator of a request is RM_CHANGE or RM_FORCED_NOTIFICATION. 9. If a registrar (possibly the same registrar) catches the connect request by selecting on the listen port in the main loop, it will errantly try to read data from the listen port. This causes the registrar to summarily close the listen port without realizing it. Consequently, all subsequent attempts to contact monitors will fail. SR: 4701420828 4701427609 4701428573 4701428672 4701428870 4701429050 8606102970 Patch Files: EMS-Core.EMS-CORE,fr=A.03.00,fa=HP-UX_B.11.00_32/64,v=HP: EMS-Core.EMS-CORE,fr=A.03.00.01,fa=HP-UX_B.11.00_32/64,v=HP: EMS-JCore.EMS-CORE,fr=A.03.00,fa=HP-UX_B.11.00_32/64,v=HP: EMS-JCore.EMS-CORE,fr=A.03.00.01,fa=HP-UX_B.11.00_32/64, v=HP: /etc/opt/resmon/lbin/p_client /etc/opt/resmon/lbin/registrar /etc/opt/resmon/lib/libopcn.1 /etc/opt/resmon/lib/libresmon.1 /etc/opt/resmon/lib/nls/msg/C/resmon.cat EMS-Config.EMS-GUI,fr=A.03.00,fa=HP-UX_B.11.00_32/64,v=HP: EMS-Config.EMS-GUI,fr=A.03.00.01,fa=HP-UX_B.11.00_32/64, v=HP: EMS-JConfig.EMS-GUI,fr=A.03.00,fa=HP-UX_B.11.00_32/64,v=HP: EMS-JConfig.EMS-GUI,fr=A.03.00.01,fa=HP-UX_B.11.00_32/64, v=HP: /opt/resmon/bin/emsui what(1) Output: EMS-Core.EMS-CORE,fr=A.03.00,fa=HP-UX_B.11.00_32/64,v=HP: /etc/opt/resmon/lbin/p_client: EMS (A.03.00 and A.03.00.01 PHSS_20684) /etc/opt/resmon/lbin/registrar: EMS (A.03.00 and A.03.00.01 PHSS_20684) /etc/opt/resmon/lib/libopcn.1: EMS (A.03.00 and A.03.00.01 PHSS_20684) /etc/opt/resmon/lib/libresmon.1: $Header: send_email_notification.c,v x.0306 98/08/05 08:35:47 $Header: log_source.c,v x.0305 98/07/29 11:00:47 hmg r Exp $ EMS (A.03.00 and A.03.00.01 PHSS_20684) /etc/opt/resmon/lib/nls/msg/C/resmon.cat: None EMS-Config.EMS-GUI,fr=A.03.00,fa=HP-UX_B.11.00_32/64,v=HP: /opt/resmon/bin/emsui: EMS (A.03.00 and A.03.00.01 PHSS_20684) cksum(1) Output: EMS-Core.EMS-CORE,fr=A.03.00,fa=HP-UX_B.11.00_32/64,v=HP: 1274924971 77824 /etc/opt/resmon/lbin/p_client 1943574588 122880 /etc/opt/resmon/lbin/registrar 4091903081 28672 /etc/opt/resmon/lib/libopcn.1 2827085420 512000 /etc/opt/resmon/lib/libresmon.1 1356537907 73795 /etc/opt/resmon/lib/nls/msg/C/resmon.cat EMS-Config.EMS-GUI,fr=A.03.00,fa=HP-UX_B.11.00_32/64,v=HP: 4049088266 135168 /opt/resmon/bin/emsui Patch Conflicts: None Patch Dependencies: None Hardware Dependencies: None Other Dependencies: None Supersedes: PHSS_18436 PHSS_18435 Equivalent Patches: PHSS_20685: s700: 10.20 s800: 10.20 Patch Package Size: 960 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_20684 5. Run swinstall to install the patch: swinstall -x autoreboot=true -x patch_match_target=true \ -s /tmp/PHSS_20684.depot By default swinstall will archive the original software in /var/adm/sw/save/PHSS_20684. If you do not wish to retain a copy of the original software, use the patch_save_files option: swinstall -x autoreboot=true -x patch_match_target=true \ -x patch_save_files=false -s /tmp/PHSS_20684.depot 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_20684.text file is available in the product readme: swlist -l product -a readme -d @ /tmp/PHSS_20684.depot To put this patch on a magnetic tape and install from the tape drive, use the command: dd if=/tmp/PHSS_20684.depot of=/dev/rmt/0m bs=2k Special Installation Instructions: Various EMS shared libraries have changed and in order for these changes to take affect, any EMS monitors or EMS clients that are running when the patch is installed will need to be stopped (kill -9) and restarted. Please refer to specific client and monitor documentation on how to restart these processes.