Messages similar to the following are displayed on the VOM console:
"The vxdclid process is not running on this host"
On the host, however ps -ef and the output of the process itself show it is running:
For the vxdclid process, “/opt/VRTSsfmh/bin/vxadm service status” shows RUNNING.
# ps -ef | grep vxdclid
root 8455 1 0 Mar14 ? 00:25:37 /opt/VRTSsfmh/bin/vxdclid
"The xprtld process is not running on this host"
# ps -ef | grep xprtld
root 7572 1 0 10:32 ? 00:00:00 /opt/VRTSsfmh/bin/xprtld -X 1 /etc/opt/VRTSsfmh/xprtld.conf
The issue happens when the SF family discovery reports the fault for dcli down condition to the VOM DB (via the dbsync call) but terminates before it could do the rest of the processing.
Once dbsync returns success the discovery process writes the reported status of the fault (and all data that was discovered) to local files. This is how it records what was reported last by this family.
In the next discovery cycle SF discovery detects dcli is up but since it did not record that it had reported a fault for this condition previously it does not report a delete for the fault.
This is the reason the fault does not get cleared from the VOM UI.
The workaround for correcting this discrepancy is to run the following:
# /opt/VRTSsfmh/bin/mh_ctl.pl -rescan 2> /dev/null
on the managed host to resync the databases.