Chapter 9. Monitoring Server Troubleshooting; Messages Related To The Index File Are Displayed When The Agent Fails Back To A Remote Monitoring Server; A Generic Rpc Communications Error Is Received When Issuing A Long-Running Tacmd Execute Command; Troubleshooting Monitoring Server Problems On Distributed Systems - IBM E027SLL-H - Tivoli Monitoring - PC Troubleshooting Manual

Chapter 9. Monitoring server troubleshooting

Review the Tivoli Enterprise Monitoring Server topics for help with
troubleshooting errors related to the monitoring server.
Messages related to the index file are displayed when the agent fails
back to a remote monitoring server
These messages indicate that the remote monitoring server was stopped forcefully
(for example, when it crashes), but that the database is not corrupted. The
messages help to ensure that even though the remote monitoring server stopped
unexpectedly, no loss of data occurred, and that the database has been restored
successfully.
A generic RPC communications error is received when issuing a
long-running tacmd execute command
A generic RPC communications error is received when you issue a long-running
tacmd execute command or tacmd executeAction command on an agent that is
connected to a remote monitoring server. Agents directly attached to the hub
monitoring server will not have this problem. When you run a tacmd
executecommand or tacmd executeAction command on an agent that is attached to a
remote monitoring server, and the command was issued with a the -t (timeout)
option with a timeout value greater than 600 seconds (10 minutes), the command
fails with a generic RPC communications error. The request does not incur a
network or communication error, but is actually being terminated by the hub
monitoring server when the response for the command is not returned within 600
seconds. However, the error returned to the TACMD indicates a communications
error.
The default hub monitoring server behavior to timeout long-running remote
requests is normally used to manage requests that have not returned within this
time period, and also to indicate a network outage might have occurred, or that
the remote monitoring server is down. However, this typical result is not the case
for a long-running tacmd executeAction or tacmd executecommand where the
command is still running at the endpoint, and the monitoring server is still online.
If you intend to run commands that take longer than 600 seconds, you can set the
KDS_SYNDRQ_TIMEOUT environment variable to run 60 seconds longer than the
expected time for the command to complete. KDS_SYNDRQ_TIMEOUT is a
monitoring server environment variable set only at the hub monitoring server. The
variable can be set in the service console for dynamic update or in the monitoring
server configuration file, which requires a hub recycle. This value can be set
arbitrarily high.

Troubleshooting monitoring server problems on distributed systems

The problems described in this section might occur on distributed systems. For
information about configuring theTivoli Enterprise Monitoring Server, refer to the
IBM Tivoli Monitoring Installation and Setup Guide .
© Copyright IBM Corp. 2005, 2012
153