Monitoring OneAgent and alerting if OneAgent alone is stopped and not the HOST

[deep_kumar_gopa]

Hi Recently we experienced issues with few servers where in one agent(version 1.205) goes down suddenly for no reason with log saying one agent stopped manually. We could not find any issues with host as well as it is up and running continiously its is only the one agent that is getting stopped. Is there any way weherin we could alert if something happens to one agent on host. Alerting of one agent goes down so tah we can atleast restart it immediately.

[Babar_Qayyum]

Hello @deep_kumar_gopa 

The Unmonitored availability state indicates that OneAgent isn't running on the host.

https://www.dynatrace.com/support/help/how-to-use-dynatrace/hosts/monitoring/host-availability/

Regards,

Babar

[deep_kumar_gopa]

Thanks Babar , But we have not received any alerts for one agent getting stopped. when no of servers is more , It takes some effort to find out whats is missing in UI. we need some way to identify when onegent get stopped.

Note : In above cases we have tried enabling alert on graceful shutdown as well but this didnt work as it is dependt on host and not One agent. 

[Rakan_khalaileh]

you can create an alerting profile, from settings go to alerting => add alerting profile => select the management zone that you need to monitor => add event filter => select predefined then from filter problems choose Host or monitoring unavailable, also you can choose monitoring unavailable if you need to get an alert if OneAgent isn't running on the host.

Br,

Rakan Alkhalaileh

[thasthakir]

Hi Rakan,

I tried to create an alerting profile and if we stop the one agent from server, We don't see alerts triggered.

[theharithsa · Accepted solution]

Hi @thasthakir 

Upon graceful process stop, Dynatrace will not create a problem as this will consider that you have shutdown the process. This is to reduce false alerts. 

Even still if you want to get the alert, you should go inside the process and turn on its availability monitoring then whenever OneAgent process stop you will get the alert - This is deliberately turned off by default in Dynatrace. 

[thasthakir]

Hi @theharithsa,

Can you please share any documentation? Should we add in OS Service monitoring? 

Please let us know.

[Sujit]

@theharithsa you wouldn't get any alert if OneAgent goes down, you will get alert only  like @Babar_Qayyum said Unmonitored availability state. Apart from that no alert. .

Thanks,

Sujit 

[thasthakir]

Hi Sujit,

So if one agent hangs or someone stops the one agent service, is there anyway we can be alerted ?

[Sujit]

@thasthakir No we wouldn't. We will getUnmonitored availability because the moment OneAgent goes down Dynatrace will lost their connectivity and get Unmonitored availability alert but not related to specific OneAgent. 

Thanks,

Sujit

[thasthakir]

@Sujit Thanks for your quick response.

Let me make it clear, If some user stops the one agent service or if the one agent hangs and crashes itself. we won't get an alert that one agent is down but we will still get the Unmonitored availability alert? 

Can you please confirm? 

[Sujit]

@thasthakir Yes, you are riight.

[thasthakir]

@Sujit Thank you ! How ever when we test by stopping the one agents we are not getting the Unmonitored availability alert as well. 

Can you please let me know if there is any specific setting for it ? 

Thank you!

[Sujit]

@thasthakir Did you check under the Problem even though if you have not received the alert but Problem card will generate. Please check Problem card.

Thanks,

Sujit

[thasthakir]

@Sujit Checked the problems, We don't see an alert when one agent is stopped from the server by an user.

[Sujit]

@thasthakir Can you run below command (if its linux) on the same host where you stopped user stopped oneagent...

service oneagent status

Or

Setting -> Monitoring Overview -> Enter the host name and see whether its Monitored or Unmonitored. 

Thanks,

Sujit

Thanks,

Sujit

[theharithsa]

Hi @Sujit, 
I was talking about the process availability monitoring. It is possible to get the alert when OneAgent is down, 

Please refer the below screenshot, 

I hope this will answer the question. 

[Sujit]

@theharithsa Ahh,... i didn't noticed this part. Thanks for providing this valuable information. 

Thanks,

Sujit

[ben_wrightson]

Have you seen my suggestion in https://community.dynatrace.com/t5/Alerting/Alert-with-Unmonitored-Host/m-p/118676 ? Would that help in your case?