It's all about the answers!

Ask a question

Starting/Stopping Jazz Applications in Websphere hangs after update to 6.0.3 M4


Philipp Waldinger (1313) | asked Aug 09 '16, 8:50 a.m.
edited Aug 30 '16, 4:13 a.m.
We use a small test system which only had JTS and RM (both release 6.0.2)  installed until the end of last week.

Then, I upgraded those two applications to 6.0.3 M4 and also installed CCM, RQM, GC, LDX, RELM and the reporting applications DCC, LQE and JRS.
Everything seemed fine, with one exception: When anybody tried to access /dcc, the following error appeared and was displayed "fullscreen" as simple text on a blank page:
Error 503: CRJAZ1173W The com.ibm.rational.datacollection.service.internal.web.IWebRedirector service is not available.
I also noticed that DCC was not found under jts/setup any more.
If I tried to add it again, I got the error that this application was already registered and the discovery url has to be uniqe.

Then I simply wanted to stop all applications and restart websphere.
So I selected all applications and presserd "Stop" in the Websphere Administrative Console.
But it took hours (literally) and after some time, I decided to simply kill the java process with the taskmanager (this was the common process for "restarting" websphere in our organization all the time before I began to maintain the server).
After websphere was then started again (with startServer.bat), all systems seemed to be running again, but the DCC error did still persists.

I also wanted to know whether I could now start/stop applications properly again (as some time before the update) and tried to "stop" all applications again. Now, over 12 hours later, the operation is still not finished and it still says "Please Wait...".

During the last days, I deleted all content in the folders "temp" and "wstemp" in the profile folder. I dont know whether this is related to the problem.

In the SystemOut.log, I find the following entry tons of times:
[09.08.16 14:44:40:239 CEST] 000001a9 WebContainer  E com.ibm.ws.webcontainer.internal.WebContainer handleRequest SRVE0255E: A WebGroup/Virtual Host to handle /ccm/oslc/workitems/trs has not been defined.
I want to know why this happens and how to solve it. Unfortunately, I'm not really a Websphere Expert (in fact, I'm quite new to it), so I don't really know where to start.
Have you ever experienced (and solved) the same behavior? What can I try to narrow the problem down?

One answer



permanent link
Ralph Schoon (63.6k33647) | answered Aug 09 '16, 10:28 a.m.
FORUM ADMINISTRATOR / FORUM MODERATOR / JAZZ DEVELOPER
edited Aug 09 '16, 10:29 a.m.
You are aware that you are working with a development milestone that is not for production and can have issues, right?
You can create a defect for the issue here: https://jazz.net/jazz/web/projects/Rational%20Team%20Concert#action=com.ibm.team.workitem.viewWelcome if you are willing to provide log data.

Comments
Philipp Waldinger commented Aug 09 '16, 10:37 a.m.

Hi Ralph,


thank you for your answer!

Yes, I'm completely aware of this. We are talking about a small testsystem (only 2 users).

But as I wrote, I dont have much experience with Websphere and the configuration of the Jazz Applications, so at first I assume its a problem related to my configuration or things I missed, not a real defect.

But ok, if you tell me it probably is a defect, I can treat it as one.

Best regards
Philipp


Ralph Schoon commented Aug 09 '16, 10:51 a.m. | edited Aug 09 '16, 10:52 a.m.
FORUM ADMINISTRATOR / FORUM MODERATOR / JAZZ DEVELOPER

Based on what information should anyone help or judge if it is your configuration or a defect? The error message could come from a lot of causes, I guess.

I know from some workshops we prepared in the past that sometimes WAS was a bit slow to shut down and I have seen that it sometimes failed to shut down. I have worked with Tomcat and WAS Liberty more and it was more responsive. I had to kill Tomcat sometimes. I don't see if WAS liberty hangs. But I am not running a big deployment and can't really learn about such an environment at the moment.

Where I would not be surprised about issues is a small deployment with only few CPU's and everything on one or two machines with potentially too few resources.

Your answer


Register or to post your answer.


Dashboards and work items are no longer publicly available, so some links may be invalid. We now provide similar information through other means. Learn more here.