ISE application server restarts randomly

[mikeyasg]

Hi, 

We have a two node ISE deployment and the primary has admin and policy and the secondary has monitoring and policy services. recently after we have upgraded to ISE 3.1 from 3.0 we are experiencing random application server restart on the primary server.  RADIUS authentications can continue because the secondary has policy services but we can not administer ISE if the primary is not available. the it happens like 5 times a day or more. application server restart will take like 15 to 20 minutes. any idea on the issue if you had some experience on this.

Thank You in advance for the help.

[balaji.bandi]

First i would check any disk space issue ? and is this VM or Appliance ?

as mentioned in the image, check the Logs and see any obnormal.

Other hand, if you have contract raise an TAC also advised here.

[ahollifield]

This is not a supported ISE configuration.  In a two-node/HA deployment you need:

ISE1: Admin  + Monitor + PSN

ISE2: Admin + Monitor + PSN

It is also best practice to split the admin/monitor primary/secondary across the two nodes.  

https://www.cisco.com/c/en/us/td/docs/security/ise/performance_and_scalability/b_ise_perf_and_scale.html

[Aref Alsouqi]

As @ahollifield mentioned you need to add the secondary admin and monitoring personas. However, I don't believe that could cause the application restart issue, it could be due to a software bug. I would try to look at the system logs to see if there is anything obv that could point to the root cause of this by using the command "show logging system ade/ADE.log" from CLI. 

[Marcelo Morais]

Hi @mikeyasg ,

 did you apply P3 (ISE 3.1 P3)

 Please take a look at: CSCwb29140 Threads getting exhuast post moving to latest patches were nss rpm is updated(Only 3.0p5&2.7p7,3.1P1.

Hope this helps !!!

[thomas]

Please call TAC for something critical like this!

This is not normal or good!

You have not provided any logs that might allow the community to troubleshoot the real issue.

See How to Ask The Community for Help for future troubleshooting requests.