Running Jira using all system cpu resources

[Samuel Hapák]

I have just installed Jira on Debian 64bit linux server and it constantly uses 100% of both of my cpu cores. Jira is freshly installed, nobody uses it at the moment, but it still uses 100% of both of my cpu cores. I have no clue, were is the problem. So I just cannot let the Jira run on the server, because it has also other responsibilities. Any clue, where is the problem?

[Samuel Hapák · Accepted solution]

I have found the solution in this article http://blog.wpkg.org/2012/07/01/java-leap-second-bug-30-june-1-july-2012-fix/

There is a great chance, this is caused by the "leap second kernel bug". Firstly, check for the following in the `dmesg`

[10703552.860274] Clock: inserting leap second 23:59:60 UTC

To fix it, firstly stop the ntp client. On debian-like systems

/etc/init.d/ntp stop

Store the current time

date -s now

And test the java. If everything is working correctly, try to restart ntp service

/etc/init.d/ntp start

And test it again.

[Azwandi Mohd Aris]

In such scenario, I'd advise you to generate a thread dump of the JVM. Here is a guideline on how to do so - https://confluence.atlassian.com/display/JIRA/Generating+a+Thread+Dump.

You can then use the lightweight Thread Dump Analyzer (also mentioned in the documentation) to inspect the threads and look for clues. If you need help, you can always post your thread dumps here (it does not contain confidential information as far as I can see) or raise a support ticket with Atlassian.