Agent not collecting data - Can we create alert for it..?

Here is the trigger condition for our alert if the agent is not responsive.

The default alert may be enough for what you are looking for. Copy "Node not polled in last 10 minutes" and add in Agent.

Hello stevenstadel

Thanks for the response. But the problem I'm facing is, node is absolutely behaving well & so the agent as well. But the agent isn't collecting any metrics for CPU or Memory. 

I think, we may need to go via some scripts which will fetch details directly from DB.

-SOSP

Hello stevenstadel

Thanks for the response. But the problem I'm facing is, node is absolutely behaving well & so the agent as well. But the agent isn't collecting any metrics for CPU or Memory. 

I think, we may need to go via some scripts which will fetch details directly from DB.

-SOSP

HI Solaiy, This example Custom SWQL alert will trigger if the LastSystemUptimePollUTC is not updated in the last 120 minutes.

SELECT Nodes.Uri, Nodes.DisplayName FROM Orion.Nodes AS Nodes

LEFT JOIN Orion.NodeSettings ns ON Nodes.NodeID = ns.NodeID AND SettingName like '%Credential%'      
LEFT JOIN Orion.Credential c ON ns.SettingValue = c.ID      
JOIN Orion.Engines e ON e.Engineid=Nodes.Engineid  
--WHERE Status<>'2'  
-- Status Not Down, UnManaged, Unreachable  
WHERE Status NOT IN ('2', '9', '12')  
AND ObjectSubType='AGENT'  
AND MinuteDiff(LastSystemUptimePollUTC,GETUTCDATE())>120 

ORDER BY LastSystemUptimePollUTC

Hello stevenstadel

Thanks much for your reply & the script. But it's kind of complex here. If we are polling for LastSystemUptimePollUTC, it will check for the polling information or heartbeat or boot time kind of, if I'm not wrong. 

But here, everything is perfect, even the heart beat of the server is perfect. But the CPU metric is not collected, likewise memory metric & volume metric is not being collected, even-though the server/agent is UP & running.

Hope you got some idea of this complex situation. 

Thanks in advance. 

We've all had this problem @solaiy, Steven's solution is a solid one for exactly what you're describing

stevenstadeladam.beedell
Sorry for the late reply here. Think this uptime UTC is working. But still I need to do some deep dive testing on all metrics for a particular machine. I'll keep you posted on this. 

Thanks much..!

-Solaiy.