Login with a specific user only on one of nodes raises the CPU

[Massy]

We have a JIRA DataCenter with 3 nodes. I have an important user on it. When I login on one of this nodes leads to CPU loading near 60%! It happens just on one node and doesn't happen for two other nodes, it takes time about 10m.

[Daniel Ebers]

Hi Massy,

this scenario happened to me with a user that had a faulty browser plugin installed. It was requesting something from the server hundred times a second - while it should not do that.

To dig deeper it would be needed to understand how your setup is structured but even then it is hard to debug from remote.

Is there any chance you log in for a test with this specific trouble user from a different client?

In case this does not help please provide more details so somebody from here is more likely being able to help.

Cheers,
Daniel

[Massy]

No. I can't.

I have these logs:

08:43:15,389 DEBUG service.batch.ScheduledReportsBatchService:79 - Total subscriptions: 0, Executed subscriptions: 0
22-Nov-2020 08:44:02.440 WARNING [https-jsse-nio-8443-exec-15] com.sun.jersey.spi.container.servlet.WebComponent.filterFormParameters A servlet request, to the URI https://X.X.X.X:8443/rest/issueNav/latest/preferredSearchLayout, contains form parameters in the request body but the request body has been consumed by the servlet or a servlet filter accessing the request parameters. Only resource methods using @FormParam will work as expected. Resource methods consuming the request body by other means will not work as expected.
08:45:11,043 DEBUG app.service.CachingServiceImpl:76 - expiring 0 keys. total 0 keys
08:45:11,043 DEBUG app.service.CachingServiceImpl:78 - expired 0 keys. remaining 0 keys
08:45:24,282 DEBUG service.batch.ScheduledReportsBatchService:79 - Total subscriptions: 0, Executed subscriptions: 0
(END)

[Daniel Ebers]

From those logs nothing suspicious is to be seen.

[Massy]

I tried to check the problem in laboratory environment. So I made a clone of troubled instance and add it to a test JIRA data-center with a similar database and the problem happened again. I tried to test adding new instance (not clone, a new instance and new installation) to the same database but it happened again.

Now, I guess the problem is from database about third instance but I don't know what it is and how to be sure about that and how to fix it! 

[Daniel Ebers]

Do you see spikes from Database monitoring (https://confluence.atlassian.com/adminjiraserver/monitoring-database-connection-usage-938847726.html) and or from SQL logs (https://confluence.atlassian.com/jirakb/enable-sql-query-logging-in-jira-server-665224933.html) when this happens?