BFConnector engine in warning state after project change
Hi there,
I have been trying to integrate Build Forge 7.1.2 (which I manage) with a RTC 2.0 server (which another IT group manages). We have been running into a lot of issues with the integration, only until recently that I have noticed that the issues may be caused by something I did with the build definition. I am hoping if anyone can help see whether this is a known issue, and if now, how I would go about resolving the issue.
The test project that I initially used was a simple one that sleeps for 30s and echoes a message before and after the .sleep command. I was able to start and cancel the build in both BF and RTC client, and the results would show up in the RTC client appropriately. However, if I change the RTC build definition to point to another project (which is very different from the simple project) and then request for a build, the job will stay in Pending state as shown in the RTC client. At this time, the RationalBuildForgeConnector will remain idle. If I cancel the build request, the RationalBuildForgeConnector engine will go from idle to warning shortly after, and things would work until the RTC server is restarted.
Here is the software being used:
Rational Build Forge 7.1.2
Rational Team Concert 2.0.0.2 iFix 3 (both server and client), with connector and client components from BF 7.1.2 installed
There is a dedicated user in both RTC and BF to manage the automated builds.
It would be greatly appreciated if someone can shed some light to the problem. Please let me know if more information is needed.
I have been trying to integrate Build Forge 7.1.2 (which I manage) with a RTC 2.0 server (which another IT group manages). We have been running into a lot of issues with the integration, only until recently that I have noticed that the issues may be caused by something I did with the build definition. I am hoping if anyone can help see whether this is a known issue, and if now, how I would go about resolving the issue.
The test project that I initially used was a simple one that sleeps for 30s and echoes a message before and after the .sleep command. I was able to start and cancel the build in both BF and RTC client, and the results would show up in the RTC client appropriately. However, if I change the RTC build definition to point to another project (which is very different from the simple project) and then request for a build, the job will stay in Pending state as shown in the RTC client. At this time, the RationalBuildForgeConnector will remain idle. If I cancel the build request, the RationalBuildForgeConnector engine will go from idle to warning shortly after, and things would work until the RTC server is restarted.
Here is the software being used:
Rational Build Forge 7.1.2
Rational Team Concert 2.0.0.2 iFix 3 (both server and client), with connector and client components from BF 7.1.2 installed
There is a dedicated user in both RTC and BF to manage the automated builds.
It would be greatly appreciated if someone can shed some light to the problem. Please let me know if more information is needed.
One answer
The RTC server admin and I were able to find out that the problem is actually an access issue. According the BF logs, the frequency of the dedicated build user logging in is higher than logging off. This seems to have caused an "access denied" error from the BF-RTC connector. Changing the event poll delay from 15s to 20s seems to work. I will close the discussion in this thread and create another thread on the actual error with the event poller. Thanks to anyone who may have helped look into this.