It's all about the answers!

Ask a question
QuestionsTagsUsersBadges
Guidelines | FAQ
Follow this question
Question details

×42,837

question asked: Jun 10 '11, 6:26 p.m.
question was seen: 5,485 times
last updated: Jun 10 '11, 6:26 p.m.

Related questions

RTC 3.0 iFix1 - Tomcat Heap full after 2 days. Is that OK?

0
Larry McCarthy (210●22●24) | asked Jun 10 '11, 6:26 p.m.
Setting the scene:
jts and ccm on out-of-the-box Tomcat, on RHEL 5.6 w/8GB RAM. jts and ccm are all it does.
Database is DB2 on a nearby IBM i.

After 2-3 days, Server status looks like so:
http://dl.dropbox.com/u/3169699/FullHeap.jpg

The logs are clean, but this rapid filling up of the JVM VM to the brim looks ... alarming. I've seen several installations of 3.0 in differnet configs. They generally lope along for days, weeks or months at something like 23% of 333MB of 4GB. I don't recall ever seeing this full VM thing happen.

IBM Support suggests this is just normal query caching and is nothing to worry about.

Should I worry?

Thanks,
Larry.

5 answers

Most liked answers ↑|Newest answers|Oldest answers

0
permanent link
Anthony Kesterton (7.5k●9●180●136) | answered Jun 12 '11, 8:58 a.m.
JAZZ DEVELOPER
Setting the scene:
jts and ccm on out-of-the-box Tomcat, on RHEL 5.6 w/8GB RAM. jts and ccm are all it does.
Database is DB2 on a nearby IBM i.

After 2-3 days, Server status looks like so:
http://dl.dropbox.com/u/3169699/FullHeap.jpg

The logs are clean, but this rapid filling up of the JVM VM to the brim looks ... alarming. I've seen several installations of 3.0 in differnet configs. They generally lope along for days, weeks or months at something like 23% of 333MB of 4GB. I don't recall ever seeing this full VM thing happen.

IBM Support suggests this is just normal query caching and is nothing to worry about.

Should I worry?

Thanks,
Larry.


Something to keep an eye on. Is there anything different about this install? How about any builds going on, or reports. Having lots of memory used is good (system should be faster as there is less disk access).

anthony
0
permanent link
Larry McCarthy (210●22●24) | answered Jun 13 '11, 10:44 a.m.
Yeah, it crashed over the weekend with out-of-memory errors in the JTS log. Good thing the client didn't close the PMR....
0
permanent link
Larry McCarthy (210●22●24) | answered Jun 13 '11, 5:15 p.m.
Sorry, Anthony - to answer your question, there were a few, small builds (2-3 source artifacts identified in change sets) going on. No dependency builds, nothing fancy, only 4 users, maybe 10 Work Items created and workflow'ed in the 3 days or so that it lasted before the OOM condition killed it.

- Larry.
0
permanent link
Anthony Kesterton (7.5k●9●180●136) | answered Jun 13 '11, 5:33 p.m.
JAZZ DEVELOPER
Sorry, Anthony - to answer your question, there were a few, small builds (2-3 source artifacts identified in change sets) going on. No dependency builds, nothing fancy, only 4 users, maybe 10 Work Items created and workflow'ed in the 3 days or so that it lasted before the OOM condition killed it.

- Larry.


Ouch - not good. Sounds like there is something going on that is chewing up memory. I would start looking for some error event that is clogging up the system (an error generated every few ms could flood memory). Definitely check to see if there are any other plugins on the RTC server. I have seen a problem with the Build Forge integration cause problems - are you using BF perhaps?

anthony
0
permanent link
Larry McCarthy (210●22●24) | answered Jun 13 '11, 5:52 p.m.
Nope, no plugins. Some small process customizations - a simpler Simple, adding one custom WI attribute and hiding most of the rest in WI editors. Not totally vanilla, but pretty close.

IBM Support has kicked the PMR up to L3. I'll let you know what happens.

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.