What AppManager for Unix Agent 6.5 issues are addressed in Hotfix 55939? (NETIQKB55939)

  • 7755939
  • 02-Feb-2007
  • 24-Sep-2007

Environment

AppManager 6.0.2
AppManager for Unix 6.5

Situation

What is the purpose of Hotfix 55939?

Resolution

Who should install this Hotfix?

This hotfix is recommended for all users running AppManager version 6.0.2 and using AppManager UNIX agent 6.5 on all UNIX platforms (except some platform specific issues).

This hotfix resolves the following issues:

  • When using the -H parameter, the UNIX agent does not bind to a specific IP address. With this hotfix, the agent now retains the IP address.
  • The UNIX agent does not failover from a primary management server to a secondary management server unless the primary management server is not accessible. With this hotfix, the failovers based on the management server's service port availability rather than the network accessiblity.
  • The performance of the UNIX agent degrades when multiple jobs are dropped on the agent using Knowledge Script Groups. With this hotfix, the performance issue has been resolved.
  • For jobs configured with Generate a new event when original event condition no longer exists option, the UNIX agent fails to generate an Event Down action for a corresponding Knowledge Script State Change event. With this hotfix, the agent generates the Event Down action.
  • The UNIX agent places the core dumps in the UnixAgent/ directory instead of the UnixAgent/bin/ directory. This means that the NetiqDiag does not collect the core dumps. With this hotfix, the NetiqDiag places the core dumps in the appropriate directory.
  • The UNIX agent fails to heartbeat the management servers which affects the UNIX agent and the management server logs. This is due to longer Machine Poll Interval in the Operator Console.
  • This hotfix resolves the heartbeat failover scenario.
  • The MemShortage Knowledge Script monitors the physical memory of a system. If the script returns a zero (0) value for the paging scan rate, it implies sufficient physical memory availability whereas a progressively higher positive scan rate (>0) implies higher physical memory usage (memory bottleneck). This hotfix resolves an error that caused the MemShortage Knowledge Script to occasionally return negative scan rate values.
  • The UNIX agent with primary and secondary management servers listening on non-standard ports, switches ports and communicates with the management servers on the default port 9001.
  • This hotfix modifies the NqmComms.xml file by creating or omitting entries wherever necessary in the configuration file.
  • During a job run, an unexpected file IO behavior results in a Perl runtime error. With this hotfix, the Perl runtime error occuring during a job run is resolved.
  • The MemShortage Knowledge Script collects swap memory statistics instead of physical memory statistics. With this hotfix, the MemShortage Knowledge Script collects the physical memory statistics.
  • The Discovery_ApacheUNIX fails to get the configuration file path during discovery. Use the Path for the Apache configuration files (semicolon-separated, no spaces) property on the Values tab to specify the path for the Apache configuration files. With this hotfix, the Discovery_ApacheUNIX get the configuration file path during discovery.
  • The Discovery_WebSphereMQUNIX Knowledge Script reports a similar error message when libmqm.so, MQDiscovery, or both files are not found. The error message report must display information related to the missing file. With this hotfix, the Discovery_WebSphereMQUNIX Knowledge Script reports with details of the missing file.
  • The maximum number of active Perl engines cannot be controlled with -N and -J options. These options cannot be used together as they are mutually exclusive. Specify the -J option when starting the UNIX agent, otherwise, the default -N NumOfPerlEngines option must be used.
  • The Measurement Interval parameter in the LogicalDiskIO Knowledge Script captures the time elapsed during data read/write .
    and transfer operations on IBM AIX systems. With this hotfix, the intermittent failure of the LogicalDiskIO Knowledge Script to accurately capture the Measurement Interval value on IBM AIX systems has been resolved.
  • The TopCpuProcs Knowledge Script generates no data or error when trying to get the entries from all processes. With this hotfix, the TopCpuProcs Knowledge Script generates data and the appropriate error messages. (IBM AIX only)
  • The ASCIILog Knowledge Script causes memory leak. Use the Maximum number of Log files to keep property in the Values tab to specify the maximum number of log files allowed during logging ASCII log entries.
  • This hotfix resolves the memory leak caused by the ASCIILog Knowledge Script. (SUSE Linux only)
  • The UNIX agent logs an shmat error and causes a core dump. This hotfix resolves the shmat error which causes a core dump. (HP-UX only)
  • The MemUtil Knowledge Script collects incorrect statistics for memory utilization (free and used physical memory, virtual memory, and swap space). With this hotfix, the MemUtil Knowledge Script collects correct statistics for memory utilization. (HP-UX only)
  • The CPULoaded Knowledge Script gathers incorrect data. Use the CpuUtil Knowledge Script to collect data for CPU's % User utilization, % System utilization, % Wait, % Idle, and Total utilization. (HP-UX only)
  • The NetInterfacesErrors Knowledge Script cannot find input error counters on bond network interface cards setup with no IP address assigned to them.
  • With this hotfix, the NetInterfacesErrors Knowledge Script finds input error counters on bond network interface cards setup with no IP address assigned to them. (Linux only)
  • The LogicalDiskIO Knowledge Script fails to monitor file-based partitions on the Linux operating system. With this hotfix, the LogicalDiskIO26 Knowledge Script monitors logical disk IO activity, including disk transfers, block reads, and block writes per second on the Linux kernel 2.6 operating systems. (Linux only)
.


fix

Installing the AppManager UNIX Agent Version 6.5 Hotfix

You must install this hotfix on agent computers, resository computers, and management servers only once.

The AM602_UNIX_65_Hotfix55939.exe download package contains the following files:

  • AM602_UNIX_65_Hotfix55939.tar
  • AM602_UNIX_65_Hotfix55939_Setup.exe
  • AM602_UNIX_65_Readme55939.htm


Note: Before applying this hotfix on HP-UX platform, take a backup of the /etc/syslog.conf file.


To install this hotfix:

  1. Download the AM602_UNIX_65_Hotfix55939.exe installation package.  https://download.netiq.com/KnowledgeDepot/AMU/Hotfixes/Recommended/AM602_UNIX_65_HotFix55939.exe 
  2. Extract the contents of the downloaded package to a specific folder. The default location is C:\Program Files\NetIQ.
  3. Run AM602_UNIX_65_Hotfix55939_Setup.exe on all repository computers and management servers to install the Knowledge Scripts, reports, and object type information. For repositories running in a clustered environment, run AM602_UNIX_65_Hotfix55939_Setup.exe on all nodes in the cluster.
  4. If you are installing the hotfix on IBM AIX, install the service pack AM602_UNIX_65_ServicePack52117.exe before running this hotfix. If you apply the hotfix without installing the service pack for IBM AIX, your system may malfunction.
  5. Log on to your UNIX computer with root privileges.
  6. Copy the extracted AM602_UNIX_65_Hotfix55939.tar file to your UNIX agent computer.
  7. In the directory where you copied the AM602_UNIX_65_Hotfix55939.tar file, enter the following command:
    tar -xvf AM602_UNIX_65_Hotfix55939.tar
    This command unpacks the files AM602_UNIX_65_Hotfix55939.sh and other platform specific library files into the current directory.
  8. Run the AM602_UNIX_65_Hotfix55939.sh shell script, to apply the hotfix. The script deploys all files needed for the specific UNIX operating system.

OR
 
If you want to run the setup by updating the server information then, run the following command:
 
AM602_UNIX_65_Hotfix55939.sh --primary=host:port --backup=host:port
 
This will installs the hotfix and modifies the NqmComms.xml by creating/omitting <SERVER></SERVER> entries wherever necessary in the configuration file as follows:
 
Usage examples:

  • Names the primary and seconday ports that will communicate on port 9001:
    AM602_UNIX_65_Hotfix55939.sh --primary=Foo:9001 --backup=Bar:9001
  • Names the primary (use whatever port it is already using) port and sets the secondary port to MS Bar on port 9001:
    AM602_UNIX_65_Hotfix55939.sh --primary=Foo:@ --backup=Bar:9001
  • Inherits the primary MS name, but sets its port to 9001. Repeats the same for backup, but replaces the port to 9030:
    AM602_UNIX_65_Hotfix55939.sh --primary=@:9001 --backup=@:9030
  • Removes the backup MS from the communication file:
    AM602_UNIX_65_Hotfix55939.sh --primary=Foo:9001 --backup=null
  • Removes the backup MS from the second communication file:
    AM602_UNIX_65_Hotfix55939.sh --primary=Foo:9001 --backup=null --config=2

The MAINT.LOG file lists all changes and problems encountered during the installation process. The MAINT.ERR log file lists only the problems encountered. By default, these file are located in the \NetIQ\Temp\NetIQ_Debug folder.



fix
This hotfix addresses intermittent crashes within the management server service when collecting data, the package AM602_UNIX_65_Hotfix55939_Setup.exe   should be installed on each management server. The management server service netiqms.exe is now build number 6.0.58745.440

note

If you are running AppManager for Unix Agent 6.0.2, please refer to the following Knowledge Base article:

NETIQKB55938:What AppManager for Unix Agent 6.0.2 issues are addressed in Hotfix 55938?



Additional Information

Formerly known as NETIQKB55939