Skip to main content
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index] [List Home]
Re: [jakartaee-tck-dev] Weekly building of Jakarta Platform Stand-alone TCKs...


On Sat, May 30, 2020 at 3:34 PM Alwin Joseph <alwin.joseph@xxxxxxxxxx> wrote:

Hi Scott,

Yes this issue is consistently occurring for some time where the job fails at the last publish-junit-report stage after all the bundles are copied.

I could not find how we can change the default injectedEnvVars for this job. I am only familiar with changing the variables for the pipeline jobs at the Jenkinsfile.

Could this issue be related to the commit : https://github.com/eclipse-ee4j/jakartaee-tck/commit/00e67287cfbfee273942bfd29b7c41edc035e3d3.

The purpose of https://github.com/eclipse-ee4j/jakartaee-tck/commit/00e67287cfbfee273942bfd29b7c41edc035e3d3 is to increase the (JNLP) container/VM memory size to be a little bigger than what we configured the JNLP Java max heap size to be.  I think this helped avoid memory exhaustion failures doing git operations.  

I think that what had been happening, is that when the JNLP Java heap size grew over time to be near the -Xmx2048m max, the container/VM was probably thrashing and falling to its knees.  This was my assertion anyway, which wasn't based on facts, but seemed to help.  I think that we can get facts or statistics about what is going on for such problems, by creating an Eclipse CI ticket.

Anyway, back to our current memory issue.  I think that we need more memory for the Java heap and yes, we also need to check that container/VM memory size is big enough for that, as having a Java heap that is larger than available OS memory, causes thrashing. :)

Regards,
Scott
 

Regards,
Alwin

On 30/05/20 11:51 PM, Scott Marlow wrote:
Hi Alwin, all,

I was just looking at the "java.lang.OutOfMemoryError: Java heap space" failure in standalonetck-nightly-build-run-master [1] , it looks like we have a max of 256mb used for the JNLP process according to injectedEnvVars [2], could someone please remind me where we are setting the environment variables for [2]?  I'm thinking that we should go higher than -Xmx256m (512m or even 1-2g) in NLP_PROTOCOL_OPTS.

Thanks,
Scott


On Tue, May 26, 2020 at 10:58 AM Alwin Joseph <alwin.joseph@xxxxxxxxxx> wrote:
Hi Scott,

I will analyse the pending issues in the standalone TCK build and ask
for help here once we have more details. Most of them would be related
to the already created issues to remove the pruned techs.

Regards,
Alwin

On 26/05/20 8:07 PM, Scott Marlow wrote:
> Alwin also created
> https://github.com/eclipse-ee4j/jakartaee-tck/issues/293 for
> correcting the building of  standalone TCKs as well.  Alwin, do you
> need help with jakartaee-tck/issues/293?


Back to the top