How can I debug an out of memory error during Rational Asset Manager startup?
The memory usage grows until reaching the maximum heap size of 2G. The memory in the server is 8G.
Any suggestion will be appreciated.
I include an extract of the exception stack trace:
.... com.microsoft.sqlserver.jdbc.SQLServerResultSet com.ibm.ws.rsadapter.jdbc.WSJdbcResultSet com.ibm.ram.repository.core.sdo.BaseRAMSchemaAccess com.ibm.ram.repository.core.StaleCacheManager com.ibm.ram.repository.web.ws.quartz.jobs.AbstractRAMJob org.quartz.core.JobRunShell .... |
3 answers
Can I just check a few things....what version of RAM are you using, and what OS? Did the installation work before, or is this a new error when the system was working fine before.
regards anthony Comments
Jose Luis Lopez Lopez
commented Jun 13 '12, 11:30 a.m.
Hi Anthony, Windows 2003, RAM 7.5.1. It was a fully operational installation with 1000 assets 10 users. I have digged into the logfiles and observed that the JobScheduler holds 96000 jobs for PendingPolicies... Is there a way to clear the scheduler?? I have cleared all the policies from the lifecycles, there should be no new jobs in the scheduler... Thanks. |
Thanks for the update on what you are running. I am not sure how to clear these logs, but can look into this.
In the meantime, if anyone else knows how to fix this - please chime in. anthony |
The best thing would be to get a dump, and look to see what what is it that takes that memory.
If you are creating assets in a batch, make sure that policies (governor configuration) are set to run as pending ... as they will not be able to catch up.... if this occured, take a look at the STALECACHE table, it should have one entry in it. Check to see its size. if it is large (> 20Meg) ... delete this record. it keep track on pending policies. |
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.