Configure IP Monitor to use WMI
We are having issues monitoring one VM server using IP Monitor. We are considering switching from SNMP monitors to WMI monitors. The SNMP monitors are able to successfully connect about 85% of the time. It's that 15% that is driving us up a wall. There are other VM's on the same host that we are also monitoring using SNMP with no issues. They connect 100% of the time. So to work around the SNMP monitor issues on the one server we would like to try using WMI monitors. We have successfully tested remote connections using WMI. But when we configure a new monitor (using IP Monitor) it fails to connect. We think it might be an issue with credentials so we are trying to track down the proper format.
Any help with the credentials for the WMI monitor or suggestions about the issue with the SNMP monitors would be most appreciated.
Any help with the credentials for the WMI monitor or suggestions about the issue with the SNMP monitors would be most appreciated.
ASKER
During the 15% time that SNMP is not responding all services are up and running normally. The only issue is the lack of SNMP response.
Do you have any event log notices? Do you have SNMP AUTH configured such that any SNMP connection is recorded? I suspect that when there is no forthcoming response the issue is that the VM is handling other higher priority responses.
ASKER
There are no event log notices for SNMP and the service is configured to only accept SNMP packets from the IP address of the monitoring server.
Does the failure 15% of the time have a pattern. I.e. occurs at certain times. occurs consistantly within the same time period i.e. from 2pm to 5pm?
Or is it sporadic. i.e. 1 out of 7 snmp polling attempts fails consitently?
Or is it sporadic. i.e. 1 out of 7 snmp polling attempts fails consitently?
ASKER
That is the most puzzling part of the whole thing. There doesn't seem to be any pattern.
See if a ping monitor reflects the same 15% issue.
ASKER
The ping monitor are always 100% successful. The issue only affects the SNMP monitors.
Ok, can you check whether the snmp service recovery option is set to restart the service? Does the system event log reflect that SNMP was restarted?
Can you check the SNMP poller to see whether it records any errors dealing with this host i.e. an incomplete response or something like that that it can not interpret?
Unfortunately, your question deals with using WMI to monitor versus determining the cause of the SNMP issue.
See if the examples at http://www.activexperts.com/activmonitor/windowsmanagement/WMI/ are helpful in using WMI to monitor this system.
Can you check the SNMP poller to see whether it records any errors dealing with this host i.e. an incomplete response or something like that that it can not interpret?
Unfortunately, your question deals with using WMI to monitor versus determining the cause of the SNMP issue.
See if the examples at http://www.activexperts.com/activmonitor/windowsmanagement/WMI/ are helpful in using WMI to monitor this system.
ASKER
The SNMP service is configured to restart after 1 minute (3x).
If I restart the service it does add the stop / start info to the event log. But there is no matching info during the times when SNMP is not working.
As for the SNMP poller we have started monitoring the server with CACTI and it records the same gaps in SNMP connectivity.
Sorry for the question title... if we can resolve the SNMP issue that would be the preferred resolution.
If I restart the service it does add the stop / start info to the event log. But there is no matching info during the times when SNMP is not working.
As for the SNMP poller we have started monitoring the server with CACTI and it records the same gaps in SNMP connectivity.
Sorry for the question title... if we can resolve the SNMP issue that would be the preferred resolution.
It sounds more and more as through there are higher preference tasks that run on the server at the time of SNMP poll failures.
If you look at the data prior to the failure, does it seem as though network, cpu usage spikes or on an upward trend? When it recovers are the data points higher than it was prior to the 'failure' or are they on a downward trend?
If you look at the data prior to the failure, does it seem as though network, cpu usage spikes or on an upward trend? When it recovers are the data points higher than it was prior to the 'failure' or are they on a downward trend?
ASKER CERTIFIED SOLUTION
membership
This solution is only available to members.
To access this solution, you must be a member of Experts Exchange.
See if you can first develop a pattern to determine under which conditions the VM is not responding to SNMP.
Are the services the VM provides functional all the time or are they as well only functional 85% of the time.
Add a ping monitor for this VM to see whether the issue is IP/network based.