Help needed with analyzing heap dump

[Raviraj Karasulli]

hi All,

Recently, we see that our Jenkins application is restarting with OOM related errors.

we took a heap dump when the heap is utilized about 60%

We got 2 suspects with the below information.

We see the below in our regular Jenkins logs:

Jan 31, 2025 7:57:00 PM WARNING com.cloudbees.hudson.plugins.folder.computed.PeriodicFolderTrigger run
Queue refused to schedule org.jenkinsci.plugins.workflow.multibranch.WorkflowMultiBranchProject@43bf10e3[Folder/job1]
Jan 31, 2025 7:57:00 PM WARNING com.cloudbees.hudson.plugins.folder.computed.PeriodicFolderTrigger run
Queue refused to schedule org.jenkinsci.plugins.workflow.multibranch.WorkflowMultiBranchProject@4fe22f4[Folder/job2]
Jan 31, 2025 7:57:00 PM WARNING com.cloudbees.hudson.plugins.folder.computed.PeriodicFolderTrigger run
Queue refused to schedule org.jenkinsci.plugins.workflow.multibranch.WorkflowMultiBranchProject@570a25fd[Folder/job3]

we have about 100 jobs that are disabled and shown in the log almost every minute.

Could you please help analyzing the issue? I have the leak-suspects summary in zip, will this contain any proprietary data? If not, I can publish it here.

Regards,

Raviraj

[Björn Pedersen]

I suspect your kept jobs are just too big in sum (and probably one bad job is especially huge).  Jenkins  does load all jobs into memory on startup, if I am not mistaken.

First step for me  would be to check the on-disk size of <jenkins-home>/jobs,  and manually prune any overly large jobs.

Then for the future use the build-discarder plugin to discard  older builds automatically.

 

Regards,

Raviraj

[Björn Pedersen]

Björn Pedersen schrieb am Mittwoch, 5. Februar 2025 um 10:18:31 UTC+1:

ravirajk...@gmail.com schrieb am Freitag, 31. Januar 2025 um 15:33:58 UTC+1:

hi All,

Recently, we see that our Jenkins application is restarting with OOM related errors.

we took a heap dump when the heap is utilized about 60%

We got 2 suspects with the below information.

That is  3.5 GB + 1.5GB roughly. For a larger installation or very frequent job runs,  you probably just need much more memory (see jenkins docs how to configure that).

[Raviraj]

Hi All,

We have a few multibranch pipeline jobs configured with 'Gerrit code review' plugin and made it to scan for any new changes regularly.

Recently, we had disabled these jobs, even then we used to see warnings like 

"Queue refused to schedule org.jenkinsci.plugins.workflow.multibranch.WorkflowMultiBranchProject@3ed[folder/job]"

After removing the schedule [ yes, for the disabled jobs ! ] we do not see such warnings and the heap is not growing as of now [ 3 days ]

We will observe for some more time and let you know if this continues to hold good.

--
You received this message because you are subscribed to a topic in the Google Groups "Jenkins Users" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/jenkinsci-users/3JM36crCu8w/unsubscribe.
To unsubscribe from this group and all its topics, send an email to jenkinsci-use...@googlegroups.com.
To view this discussion visit https://groups.google.com/d/msgid/jenkinsci-users/1ffd385b-6ae0-4912-ac38-e3c30fd10d4cn%40googlegroups.com.

--

Regards,

Raviraj