PTC Integrity Plugin forces a stuck in Jenkins Master Server

[AlexRuff]

Hello PTC ALM Support,

in our setup PTC Plugin V2.1 and Jenkins 2.36 we face sporadic hang up of the Jenkins Master when a PTC Project is freshly checked out. My understanding is that the checkout commands are executed by the Jenkins master on the particular Jenkins node. In the error event the master gets stuck and completely unrelated Jobs currently executed get also stuck because of this.

This seems to happen with a higher chance when a new Jobs is created and the master has to check out the very first time. Once a checkout went through the job seems not to get stuck anymore.

Attached a log of Integrity PTC Plugin.

You can see that for 50min (2:33PM to 3:20PM) nothing was logged until we finally had to force restart the server at 3:20pm.

Let me know if I can provide any more logging information for further debugging.

Best regards

Alex

[apawar]

Hi Alex,

did console output show any error when the issue happened? I sense if project path has any typo or incorrect. then this issue could occur.

if the console output is contains following exception:
java.util.concurrent.ExecutionException: java.lang.Exception: Authentication failed for <Username> on <IntegrityServer>:<Port>

you can get the details in this article CS271410.

Further, could you please share console output?

Regards,

Amit

[AlexRuff]

Hello Amit,

thank you for your reply.

The Syslog of the Plugin is attached to my thread opening. This is the console output I get when Jenkins is stuck:

17:45:06 Started by user Someone
17:45:06 Building remotely on SomePC in workspace D:\BuildServer\somepc\workspace\SomeProject_compile
17:45:06 Checkout started using remote client for :#/Appl#SomeProject/AWD_AD_Slave#d=HTFBplus_FN60_ADS
17:45:06 Change Log: http://Someserver:8080/jenkins/job/Someproject_compile/4/changes
17:45:06 Build Log: http://Someserver:8080/jenkins/job/Someproject_compile/4/console
17:45:06 Preparing to execute si projectinfo for #/Appl#SomeProject/AWD_AD_Slave#d=HTFBplus_FN60_ADS
17:45:07 Preparing to execute pre-build si checkpoint for #/Appl#SomeProject/AWD_AD_Slave#d=HTFBplus_FN60_ADS
17:45:22 Successfully executed pre-build checkpoint for project #/Appl#SomeProject/AWD_AD_Slave#d=HTFBplus_FN60_ADS, new revision is 1.250.1.5
17:45:22 Preparing to execute si viewproject for #/Appl#SomeProject/AWD_AD_Slave#b=1.250.1.5

Nothing out of the ordinary. It just stuckst after printing the si viewproject command and then after 4 hours we restart the server.

I dont think the issue is an authentication or project path error, as the same configuration is working after I retry building the job a couple of times. Also the commands before viewproject are running on the same project path so it should not be invalid.

My general problem is that it is hard to obtain logs of the situation as we lose log context when we have to force restart the server because of this situation.

Best regards

Alex

[PaulG]

I am seeing a similar problem.  Seems to happen once in about 10 builds.  Any resolution to this?