| Summary: | signing broken on new hudson instance. | ||
|---|---|---|---|
| Product: | Community | Reporter: | Bouchet Stéphane <sbouchet> |
| Component: | CI-Jenkins | Assignee: | Eclipse Webmaster <webmaster> |
| Status: | RESOLVED FIXED | QA Contact: | |
| Severity: | normal | ||
| Priority: | P3 | CC: | adolfosbh, david_williams, jacek.pospychala, nicolas.bros |
| Version: | unspecified | ||
| Target Milestone: | --- | ||
| Hardware: | PC | ||
| OS: | Linux | ||
| Whiteboard: | |||
|
Description
Bouchet Stéphane
maybe this can help, this is the log from the sign action : WARNING: /opt/users/hudsonbuild/.hudson/jobs/emf-eef-integration/workspace/build/signing/I201009010445-out is not a valid directory. Creating. WARNING: outputDir /opt/users/hudsonbuild/.hudson/jobs/emf-eef-integration/workspace/build/signing/I201009010445-out not writable by user genie. Changing permissions to make file group-writable. WARNING: File /opt/users/hudsonbuild/.hudson/jobs/emf-eef-integration/workspace/build/signing/emf-eef-Master-incubation-I201009010445.zip not writable by user genie. Changing permissions to make file group-writable. File /opt/users/hudsonbuild/.hudson/jobs/emf-eef-integration/workspace/build/signing/emf-eef-Master-incubation-I201009010445.zip added to queue. You will receive notification when the file is signed, if you used the mail parameter. You can check signing status by tailing /home/data/httpd/download-staging.priv/arch/signer.log D'oh! Ok so the problem is the signer can't find the file to be signed since it's not where it's expected on the signing host(presently build). I've leveraged some code we had for build2 and I'm running an EEF build to confirm the issue is fixed. -M. Will this work automagically for hudson slave 3, 4 7 10, and 154? As long as build remains the signing node, and the slaves all mount the same job partition, yes. -M. Signing seems to be working now.
I notice that your job is still reporting failed as it looks like your forcing the pack200 path:
[build] get.baseoswsarch:
[build] [echo] Pack with /opt/public/common/ibm-java2-ppc-50/jre/bin/pack200
You can either use: /shared/common/sun-jdk1.6.0_21_x64/bin/pack200 , or just call pack200 and let the Hudson users shell deal with it(it points to the given path)
-M.
(In reply to comment #5) > Signing seems to be working now. > > I notice that your job is still reporting failed as it looks like your forcing > the pack200 path: > > [build] get.baseoswsarch: > [build] [echo] Pack with > /opt/public/common/ibm-java2-ppc-50/jre/bin/pack200 > > You can either use: /shared/common/sun-jdk1.6.0_21_x64/bin/pack200 , or just > call pack200 and let the Hudson users shell deal with it(it points to the given > path) > > -M. i do not know how to modify this, i am using the cbi build system... the pack200 path is automaticcaly set during execution, using JAVA_HOME sets to sun-jdk1.6.0_21_x64, so i don't know why it's trying with ibm-java2-ppc-50. anyway, thanks for your help for fixing the signing issue. HI, Signing fails again this morning... see https://hudson.eclipse.org/hudson/job/emf-eef-integration/79/ It looks like this was caused by the slave process having the 'wrong' group and the sign process wasn't able to correct that. I've restarted the slave and confirmed that it has the right group, and I've fired your build and it seems to be 'happy' again. -M. Hello, Reopening this bug, since I'm having similar problems in my hudson's job. checkForSignedZip: [exec] Result: 2 [echo] ls: cannot access /opt/users/hudsonbuild/workspace/cbi-mdt-ocl-3.0-integration/build/signing/I201009270432-out/mdt-ocl-Master-I201009270432.zip: No such file or directory [retry] Attempt [0]: error occurred; retrying... [echo] Sleep for 180 seconds... -timestamp: [echo] 04:44:16 Trying to override old definition of task resolvepath checkForSignedZip: [exec] Result: 2 [echo] ls: cannot access /opt/users/hudsonbuild/workspace/cbi-mdt-ocl-3.0-integration/build/signing/I201009270432-out/mdt-ocl-Master-I201009270432.zip: No such file or directory [retry] Attempt [1]: error occurred; retrying... [echo] Sleep for 180 seconds... -timestamp: [echo] 04:47:16 Trying to override old definition of task resolvepath checkForSignedZip: [exec] Result: 2 [echo] ls: cannot access /opt/users/hudsonbuild/workspace/cbi-mdt-ocl-3.0-integration/build/signing/I201009270432-out/mdt-ocl-Master-I201009270432.zip: No such file or directory [retry] Attempt [2]: error occurred; retrying... [echo] Sleep for 180 seconds... Terminating xvnc. Build was aborted Publishing Javadoc Archiving artifacts https://hudson.eclipse.org/hudson/job/cbi-mdt-ocl-3.0-integration/132/console I must note, that it's the first time I run the job in the new hudson.eclipse.org server, I mean, I've not got to make the job run in the new server. I've had other problems with the job which I've already fixed, but I'm currently having issues with the signing phase, which I can't figure out why they are occuring. As additional info, this job uses cbi Athena's build process. As far as I know, I don't have to do anything special task in my job or build configuration, and this job was working in the old build.eclipse.org server. Regards, Adolfo. > Reopening this bug, since I'm having similar problems in my hudson's job.
>
> checkForSignedZip: [exec] Result: 2 [echo] ls: cannot access
> /opt/users/hudsonbuild/workspace/cbi-mdt-ocl-3.0-integration/build/signing/I201009270432-out/mdt-ocl-Master-I201009270432.zip:
> No such file or directory [retry] Attempt [0]: error occurred; retrying...
> [echo] Sleep for 180 seconds... -timestamp: [echo] 04:44:16 Trying to override
> old definition of task resolvepath
Our Hudson instance setup has changed when the Build server decided to die. You must place files in the staging directory in the shared location for signing to work.
/home/data/httpd/download-staging.priv/modeling/mdt/ocl
The build server (which performs signing) cannot access local workspaces. I've verified the above directory exists and is accessible by your group.
I think there's nothing left to do here. Please reopen if I'm missing something. |