cancel
Showing results for 
Search instead for 
Did you mean: 
Create Post
Level 12

F5 Alert - when status is Unknown or LTM is an unknown state

we have 2 F5's in Production - when the LB admin did an upgrade this past Monday - while on his end both f5's were up and running as expected but in Orion the Primary F5 basically become a brick .. status Unknown and status reason LTM is in an unknown state and of course was not reporting on any virtual servers or Pools and Pool Members. How could I alert on this same type of scenario again?  I did look to set up a new alert for an F5 devices but didn't see any drop down choice to key off of.

0 Kudos
3 Replies
Level 15

You may need to play around with an f5 to see which particular status is the one you need to trigger on, but I *think* the Orion Status would be your best bet for the LTM Server.

2018-09-26_16-55-33.jpg

Screenshot from NPM 12.1

0 Kudos

zackm​ so this was the strange part - the node itself was up and responding to SNMP but the F5 stuff wasn't sending any info back - i had the F5 eng start with restarting the snmp service to see if that would resolve it and it did. This all happened after he upgraded the version on the f5

If you run this against your database, you can see the various sources of data for the LTM Server object:

SELECT * FROM F5_LTM_Server_V

The reason I think you need to use the "Orion Status" object, instead of "F5 Status" is precisely what you mentioned re: SNMP failing. Almost exclusively, the 'Unknown' status for all objects in Orion is reserved for one of 2 scenarios:

  • The credential being used is no longer working / The object is not responding to SNMP/WMI/Agent requests
  • The Orion object no longer exists (commonly seen if you vMotion a VM and the Volume Index number(s) change. All the "old" volumes will be seen as 'Unknown' and will need to be re-added with their new index numbers)

What I think happened to you was that after the F5 came back online, the SNMP service was in a faulted state. The next time Orion tried to poll that device, SNMP failed, which resulted in an Unknown status.

If that assumption is accurate, then creating an alert like I mentioned before, using the "Orion Status = Unknown" as a trigger, should alert you to a similar problem in the future.

0 Kudos