Can you export the log from the nsh script job run and attach it ?
Ok, so there’s a deploy job that runs that copies over the metadata from the repo and does some stuff. that’s failing most likely. If you open the NSH/patch/linuxpu/scripts/jython/linuxpatchanalysis.jli (something like that) there should be a call to a cleanup function near the bottom of the file. comment that out and then re-run the analysis. That should leave a bldeploy job in the patch analysis / scratchpad folder (something like that) and then we can see if the job ran and why it failed.
Please find the last few lines of these files, i didn't see any code related to deleting deploy Job. Also i have attached file along with this post
for server in patchesByServer.keys():
if not SharedPayloadFlag:
for patch in patchesByServer[server]:
dbkey = str(spt.getPatchKeyByName(patch))
if(dbkey != None):
_postcmd = 'sh %s/linux-deploy.sh "" 2> %s/linux-deploy.sh.output;' % (host_bl_loc, host_bl_loc)
platform = sharedPayload_obj.getServerPlatform(server)
blpkggrp, pkgname = sharedPayloadGrp, sharedPayloadPkgDict[platform]
# Save updates to software deploy job
jobCreated = 1
sys.stderr.write("Cannot create job group %s in %s \n" %(server, str(spt.getJobGroupName())+"/Deploy"))
if (jobCreated == 1):
spt.createBatchJob('Linux Patch Deploy Batch_'+uniqueName,autoExecute)
spt.createBatchJob('Linux Patch Deploy Batch_'+spt.getTimestamp())
jobList = spt.getDeployJobList()
for j in range(len(jobList)):
# Add schedule to batch job if -s is specified
# Execute batch job immediately if -e is specified
if('-s' in args.keys()): spt.addScheduleToBatchJob(args)
elif('-e' in args.keys()): spt.executeBatchJob()
if __name__ == '__main__':
loadAndDeployPatches.jli 12.0 K
in the linuxpu/Scripts/Jython/linux-analysis.py file around like 1326 there's a call to the cleanup function. comment that out and then re-run your job. then there should be a 'scratchpad' job folder left under your patch analysis job folder structure. look in there for a deploy job. get the log. and get the corresponding bldeploy log from the target.
I saw one deploy job has been created under scratchpad folder and fails with following error
There was a problem resolving the soft links in the package: java.io.IOException: JNI file copy from '/C/Program Files/BMC Software/BladeLogic/NSH/patch/linuxpu/Work/e0dc9946a8e7b84ea44c1b101e76e85b/patch-meta.tar.gz' to '//hcslu9030-gms/var/tmp/stage/fc2aec26253d370aa3aeec5f69bc790b/161133.1/patch-meta.tar.gz' failed: : No such file or directory
When i checked for file patch-meta.tar.gz, file is present at mentioned location.
Ok, does that file exist ? did you ‘yummify’ your repo w/ the offline downloader?
What options do you have set in the analysis job?
What’s in the patch/linuxpu/Work/linuxrepo.conf ?
I have attached linuxrepo.conf file and Patch analysi Job configuration also i have executed following command while repository creation
./RedHatRepositoryManager.sh -command yummifyRepo -repoLocation /Patching/CentOS/COS6x86_64/
CentOS patch analysis job is successful... but in details job logs i can see some error in .py as well .jli files. I have attached screen shot and job logs for further reference.
in the patch/jython.conf file, i think you may need to set a value for LINUX_DEPLOY_PRE_POST_CMD_STAGING_DIR=
or you need to disable the 'shared payload' mode and/or 'multi datastore' mode (set a 1 for that in the options)
I didn't see any parameter with name LINUX_DEPLOY_PRE_POST_CMD_STAGING_DIR in jython.conf file do we need to create it explicitly. I tried excluding some patches from analysis and job has been failed with error. I have added logs and screen shots for more details
did you try changing the share payload and/or multidatacenter mode settings? setting includes or excludes won't really do anything to affect the error here.
can you post the jython.conf ?
Yes, I have checked with enabling MultiDatastore mode and shared payload mode. Please find the logs for both the jobs as well as jython.conf file
well this seems like a problem ( from the shared payload ):
Run at 01/12/2015 05:32:31,Info,01/12/2015 05:48:36,Scanning host hcslu9030-gms for platform... Cannot resolve OS Release from /etc/redhat-release on host hcslu9030-gms
what's in the /etc/redhat-release file ?