bi_server1- (managed service) goes down when there is overload
Hi,
My environment is as below:
OBIEE 11.1.1.7
RHEL 5.5
RAM 50GB
There are number of analytic report created in the presentation server. We have multiple users accessing these reports. We also have huge reports being generated using BI Publisher. The heap memory provided is 20GB. When there is an overload on the server, below is what is observed:
The bi_server1 goes down with the error in the following files:
nqserver.log
[2013-04-28T10:35:02.000+03:00] [OracleBIServerComponent] [NOTIFICATION:1] [] [] [ecid: 004qqQYVv9n7Q9s6wjfd6G0000UR000000] [tid: 4400d940] Could not connect to the authentication web service (taking OBIS offline) 172.28.4.204: [nQSError: 46169] An HTTP operation to 172.28.4.204:9502 timed out after 60 seconds. Operation aborted.
biserver_1.log
<783dc1f7719d4263:-4faee90c:13e4f836034:-8000-00000000000006f1> <1367134551020> <BEA-002911> <WorkManager weblogic.kernel.Default failed to schedule a request due to java.lang.OutOfMemoryError
java.lang.OutOfMemoryError
>
####<Apr 28, 2013 10:38:35 AM AST> <Error> <Server> <etl4> <bi_server1> <DynamicListenThread[Default]> <<WLS Kernel>> <> <783dc1f7719d4263:-4faee90c:13e4f836034:-8000-00000000000006f6> <1367134715073> <BEA-002608> <The ListenThread because of an error: java.lang.OutOfMemoryError: java/net/Inet4Address, size 24B
java.lang.OutOfMemoryError: java/net/Inet4Address, size 24B
bipublisher.log
[2013-04-29T09:42:58.096+03:00] [bi_server1] [WARNING] [] [oracle.xdo] [tid: 11] [userId: <anonymous>] [ecid: 783dc1f7719d4263:3cb2800:13e519361f4:-8000-000000000000b9f7,0] [APP: bipublisher#11.1.1] oracle.xdo.XDOException: java.lang.OutOfMemoryError: allocLargeObjectOrArray: [B, size 115280856[[
at oracle.xdo.common.xml.XDOSAXParser.invokeParse(XDOSAXParser.java:216)
All the 5 OBIEE components are still alive. However, the bi_server1 (managed service) goes down. We cannot access any of our reports, or login to console or login to EM. All we need to do is restart our services and managed service to get it running again. As this is our production server, we have a lot of user complains.
Awaiting your kind feedback.
Regards,
Zuzar