Hi,
Is there any way to use APM to create a Unix FS monitoring ?
How could I create an APM agent to show me anything like this:
FS /data/devices 95% usage
Thanks
Provided SNMP is enabled and configured on the Unix host you should be able to list resources on the node and select the volumes you want to monitor as pictured below.
You could use APM's Unix/Linux script monitor parsing out the % Capacity value from a "df" command
Something like;
df -k filesystem_name_here | grep % | awk '{ print $5 }' | sed 's/%//g'
When I list resources on my Unix servers not all of the file systems show up. I am particularly concerned with some of the oracle data directories. These servers use NET-SNMP as their agent and have a very basic configuration.
aLTeReGo
If the volumes are not appearing then they may be directories or symbolic links off the root. If they appear when running "df" from the command line but not via SNMP then I would recommend upgrading the version NET-SNMP running on that server.
I'm using Linux agent to monitor one of the Linux servers using NPM. However, the / file system is not getting monitored. The data is not collected for / and when I do "list resources", I could see that the "/" is not selected. If I try to select and submit and then again do "list resources" I could see that / is still not selected for monitoring.
Could someone please help?
@aLTeReGo
@Deepthi wrote: Hi,I'm using Linux agent to monitor one of the Linux servers using NPM. However, the / file system is not getting monitored. The data is not collected for / and when I do "list resources", I could see that the "/" is not selected. If I try to select and submit and then again do "list resources" I could see that / is still not selected for monitoring.Could someone please help?@aLTeReGo
Out of curiosity, what version are you currently running in your environment? This issue appears to have been solved in the latest release under internal case DC-604.
Hi @aLTeReGo ,
I'm using NPM 12.3 and I do not see the issue with same version of NPM in another instance of Solarwinds Orion.
Agent version: swiagent-2.0.0.70-857826
Thanks,Deepthi
The issue is specific to the node being monitored via the Agent. If you were to have that other NPM instance monitor that same machine using the Agent, you would see the exact same thing you are now.
The issue has to do with how that machine is returning volume indexes to the Agent. This is a known issue that is resolved in the latest release, but is not broadly applicable to all Linux machines. In fact, this likely affects less than a single percentage of all monitored Linux machines.
Hi @aLTeReGo
Thank you for the update. So is it addressed in the latest version of NPM or agent? Is there any hotfix as I will not be able to upgrade NPM. Is there any other solution to address the issue?
Thanks,
Deepthi