IBM E027SLL-H - Tivoli Monitoring - PC, Troubleshooting Manual

Page: 197 / 310

The monitoring agent does not support the displayname/atomization, so the
SNMP alert emitter will only emit an SNMP alert for the first row in the disk table
where the situation is true. This limitation applies to subnodes as well. An instance
of a subnode agent collects metrics for all subnodes in one table. These metrics are
filtered by subnode when displayed in the Tivoli Enterprise Portal, but situations
running against multiple subnodes in an instance are running against a single
table. If a situation becomes True for 1 subnode, an SNMP alert defined for that
situation will be emitted, but no SNMP alerts will be emitted for any other
subnodes for that situation, since no further rows are processed in the table.
Binary Path attribute of the Windows OS agent does not show a value
On 64 bit systems with an installed 32 bit Windows OS agent, the value of the
Binary Path attribute of the NT_Process attribute group is null if the process is
running as 64 bit native.
Installing pre-v6.2.1 Monitoring Agent for Windows OS onto a v6.2.1 or
later monitoring server inadvertently unconfigures the monitoring
server
When a pre-v6.2.1 Monitoring Agent for Windows OS is installed onto a system
that has a v6.2.1 or later monitoring server, the agent installation inadvertently
unconfigures the monitoring server. This results in the monitoring server being left
in a stopped state. Settings for the monitoring server remain preserved in the
monitoring server ini and env files. Attempts to manually start the monitoring
server fail with the message:
Unable to start tems service. Check the event log.
Monitoring Agents for Windows OS that are v6.2.1 and higher can be installed
onto a monitoring server with no adverse side effects. IBM Tivoli Monitoring
Application Agents regardless of version can also be installed onto a monitoring
server with no adverse side effects.
If this issue is encountered, the monitoring server should be re-configured and
restarted from Manage Tivoli Enterprise Monitoring Server or through the CLI.
Settings for the monitoring server from before the installation (preserved in the
monitoring server ini and env files) will automatically be used.
OS agent restarted unexpectedly on heavily loaded systems
When Agent Management Services has a high workload, the OS agent is restarted
automatically. At this point the OS agent is stopped, and its workspace on the
Tivoli Enterprise Portal is grayed out, then the monitoring agent starts
automatically by Agent Management Services in several seconds. When the OS
agent is started, refresh workspace, you find only the Watchdog's management
status workspace as "managed", all of the other agents'workspaces' status display
"Not managed."
On a heavily loaded system, increase the command time out parameter
(KCA_CMD_TIMEOUT) in the OS agent's ini file from its default value to
something larger, up to 120 seconds. Also, you may need to increase the
checkFrequency value in the OS agent's availability checking policy file to a larger
value. See the IBM Tivoli Monitoring Administration Guide's chapter on Tivoli
Agent Management Services for information on how to change the parameters of
this file.
Chapter 10. Monitoring agent troubleshooting
179