Forums

Articles
Create
cancel
Showing results for 
Search instead for 
Did you mean: 

Do I need to change default JVM memory settings: GC (Allocation Failure)

Steve Allen
Contributor
December 10, 2018

We have just migrated to Jira 7.13.0 and everything is running fine.

Issue
I notice that I have the following entry types in: 
    /opt/atlassian/jira/logs/atlassian-jira-gc-2018-12-XX_XX-XX-XX.log.0.current
     [GC (Allocation Failure) [PSYoungGen: 239641K->4704K(249856K)] 751323K->516734K(774144K), 0.0617082 secs] [Times: user=0.09 sys=0.00, real=0.06 secs]

Further entry examples at end of details section

I have endeavoured to research similar issues and there exists good instructions on amending JVM memory values but they come with warnings as well. We are a small installation and would like advice on whether these entries really represent something I should look at tuning and what these values may be.

Hopefully the following stats will be (at least some) of the info required to help/advise:

Java VM Memory Statistics
Tools  > System.  Select System support > System Info > System Info page.
Total Memory: 755MB
Free Memory: 97MB
Used Memory: 13% Free (Total 755MB)

Jira installed within a Debian 9 VM. ESXi Resources: 4 core - 8Gb ram.
The ESXi host is a  Dell R630 64Gb ram - 32 logical processors - the host is not 'over-extended'. Jira and a similar size Confluence instance are installed within this same VM.

free -m
                    total        used        free        shared  buff/cache   available
Mem:           7987        3758        1665         148        2563        3816
Swap:          8189           0            8189

/opt/atlassian/jira/logs/catalina.out
org.apache.catalina.startup.VersionLoggerListener.log Command line argument: -Xms384m
org.apache.catalina.startup.VersionLoggerListener.log Command line argument: -Xmx768m

 

Many thanks,

SteveA 

 

atlassian-jira-gc-2018-12-XX_XX-XX-XX.log.0.current

2018-12-10T14:29:56.062+0000: 335086.206: [GC (Allocation Failure) [PSYoungGen: 239641K->4704K(249856K)] 751323K->516734K(774144K), 0.0617082 secs] [Times: user=0.09 sys=0.00, real=0.06 secs]
2018-12-10T15:38:40.122+0000: 339210.266: [GC (Allocation Failure) [PSYoungGen: 243296K->7329K(249344K)] 755326K->521417K(773632K), 0.0537629 secs] [Times: user=0.08 sys=0.00, real=0.05 secs]
2018-12-10T16:12:14.742+0000: 341224.886: [GC (Allocation Failure) [PSYoungGen: 245921K->5692K(250368K)] 760009K->521631K(774656K), 0.0484997 secs] [Times: user=0.07 sys=0.01, real=0.05 secs]
2018-12-10T16:16:35.876+0000: 341486.020: [GC (Allocation Failure) [PSYoungGen: 244284K->2411K(249344K)] 760223K->520999K(773632K), 0.0484928 secs] [Times: user=0.07 sys=0.00, real=0.05 secs]
2018-12-10T16:16:40.446+0000: 341490.590: [GC (Allocation Failure) [PSYoungGen: 241003K->10326K(249344K)] 759591K->530183K(773632K), 0.0453647 secs] [Times: user=0.08 sys=0.00, real=0.05 secs]
2018-12-10T16:16:58.465+0000: 341508.608: [GC (Allocation Failure) [PSYoungGen: 248918K->1579K(249344K)] 768775K->522297K(773632K), 0.0480268 secs] [Times: user=0.07 sys=0.00, real=0.05 secs]
2018-12-10T16:16:58.513+0000: 341508.657: [Full GC (Ergonomics) [PSYoungGen: 1579K->0K(249344K)] [ParOldGen: 520717K->476712K(524288K)] 522297K->476712K(773632K), [Metaspace: 286879K->286238K(1320960K)], 0.8932031 secs] [Times: user=2.30 sys=0.02, real=0.89 secs]
2018-12-10T16:16:59.746+0000: 341509.890: [GC (Allocation Failure) [PSYoungGen: 236544K->4192K(241152K)] 713256K->480904K(765440K), 0.0400680 secs] [Times: user=0.06 sys=0.00, real=0.04 secs]
2018-12-10T16:17:00.381+0000: 341510.524: [GC (Allocation Failure) [PSYoungGen: 240736K->1964K(248832K)] 717448K->478845K(773120K), 0.0375442 secs] [Times: user=0.05 sys=0.00, real=0.03 secs]
2018-12-10T16:17:09.697+0000: 341519.841: [GC (Allocation Failure) [PSYoungGen: 239020K->7989K(248320K)] 715901K->484906K(772608K), 0.0468330 secs] [Times: user=0.06 sys=0.00, real=0.05 secs]

1 answer

0 votes
Alexis Robert
Community Champion
December 10, 2018

Hi @Steve Allen

 

you should change your JVM maximum memory setting, the current one is set to 768m and is the default when installing Jira.

I would recommend increasing it to 1024 and see if you keep experiencing issues after that.

You can change this in your <jira_install_dir>/bin/setenv.sh and look for JVM_MAXIMUM_MEMORY= , then change it to 1024 instead of 768.

The documentation for increasing Jira JVM is worth a read: https://confluence.atlassian.com/adminjiraserver/increasing-jira-application-memory-938847654.html

 

Let me know if you have any issues, 

 

--Alexis

Steve Allen
Contributor
December 11, 2018

Thanks Alex,

I have increased JVM_,MAXIMUM_MEMORY to 1024 and confirm total memory of 996 MB.

Log  atlassian-jira-gc-2018-12-11_08-51-42.log.0.current still shows entries of type:

2018-12-11T08:56:03.953+0000: 261.347: [GC (Allocation Failure) [PSYoungGen: 296531K->21343K(317440K)] 767845K->492680K(1016832K), 0.0579815 secs] [Times: user=0.09 sys=0.01, real=0.06 secs]

I am most likely the only user logged in at this time.
I will continue to monitor.

Regards,

SteveA

Alexis Robert
Community Champion
December 11, 2018

Hi @Steve Allen

 

this might be because you had a different JVM setting before your upgrade but this change was not made on your new install.

Do you still have the old setenv.sh file from your previous version and can check what were the JVM settings in there ? 

 

--Alexis

Steve Allen
Contributor
December 11, 2018

Hi Alex,

To be clear, we migrated to a new VM (due to the updated system requirements for Jira 7.13), Jira data exported and imported etc. Followed the guides and process completed fine.


Temporarily booted up the original VM and it also has a 768m. While I was there, I checked it's log and it also has the same entry types " [GC (Allocation Failure) [PSYoungGen:".

Not sure where this leaves us :-)
Due to the migration, I was merely looking deeper for confirmation that everything was working as expected. Perhaps it is and there is no issue at all ?

Alexis Robert
Community Champion
December 11, 2018

Well, if you don't see any issue in your day to day use then it's fine, but you should try tuning the JVM to avoid those GC failure errors as it means your instance might respond slower to requests due to those GC issues.

I'd recommend having  a look at the GC tuning guide from Atlassian.

 

Let me know if it helps, 

 

--Alexis

Suggest an answer

Log in or Sign up to answer