This content has been marked as final. Show 9 replies
To answer to this question need more details about your configuration...
0) Is it new installation and configuration? whats the latest patch?
1) Does your DAC has email notifications set? if Yes did you get any email status about DAC server status? if you dont have it then I would suggest to config it.
2) What is the database server status is it up and running? In general DAC server is down when db is not available
3) Is it happening given workflow/task running or what event crash is occuring?
4) Can you let me know whats the maximum idle time you can see on DAC client with no job running?
Edited by: Srini VEERAVALLI on Jan 16, 2013 6:44 AM
0) This is not a clean install it has been running since around July 2012 Dac Version that I am running is - 18.104.22.168.2 that comes with OBIApps
1) I do not have email notifications setup I was unaware of this (my dac server runs on a linux box) I do have ETL failed/completion notifications - I did not get an ETL failed notification when dac crashed.
2) Database has not been down and has been up and running for some time.
3) No specific task is causing it to crash, to make things a little more difficult it was a few months ago was the last time this happened.
4) Can you let me know whats the maximum idle time you can see on DAC client with no job running? I am unsure what you mean here. If I dont start an ETL the dac server will idle for weeks and weeks.
No Java or Solaris updates have been done on this particular box anytime recently.
Ive increased the java heap size to 2Gb to mimic the other environment, unfortunately have no way to know or test to see if this resolves my issue.
I still think its an odd one how the logs just end, dac being based on java it can only point to an error there. Do you know of any java logs that can be looked at to shed some more light on the issue, Ive had a look around other sources but have found nothing worthwhile in relation to the issue.
Found this in nohup.out it reveals the issue is with Java, not sure how this would apply to windows to help you out maybe event viewer but I cant be sure
2013-01-15 23:13:50.818 CREATE INDEX:W_AP_INV_DIST_F:W_AP_INV_DIST_F_F23 has started.
# A fatal error has been detected by the Java Runtime Environment:
# SIGSEGV (0xb) at pc=0xffffffff7ed63284, pid=2159, tid=84
# JRE version: 6.0_26-b03
# Java VM: Java HotSpot(TM) 64-Bit Server VM (20.1-b02 mixed mode solaris-sparc compressed oops)
# Problematic frame:
# C [libc.so.1+0x63284]# [ timer expired, abort... ]
$ pldd 23497
23497: /usr/jdk/instances/jdk1.6.0/bin/sparcv9/java -server -Xms256m -Xmx800m
$ pldd 23495
23495: ksh startserver.sh
I suspect the above pldd is now outdated as I have restarted the dac server since
Edited by: DilbertsDog on 16-Jan-2013 09:18
Might be DAC 32bit running on 64bit OS
Check this link
Edited by: Srini VEERAVALLI on Jan 16, 2013 11:27 AM
Is your DAC using jdk from hits own java? I mean \orahome\10gR3_1\jdk\
if you use this DAC running on jdk 32bit
Edited by: Srini VEERAVALLI on Jan 16, 2013 11:32 AM
Just to keep you in the loop, I suspect that the crash may have been down to Java garbage collection. I have increased the initial heap size and set the young generation.
/usr/jdk/instances/jdk1.6.0/bin/sparcv9/java -server -Xmn500m -Xms2048m -Xmx204
to answer your question on java I use java that is on the box you can see the location above,
java version "1.6.0_26"
Java(TM) SE Runtime Environment (build 1.6.0_26-b03)
Java HotSpot(TM) 64-Bit Server VM (build 20.1-b02, mixed mode)
Still looking to find an exact answer to why it crashed
Going to write a script that looks for the DAC java process disappearing and if so restart DAC using startserver.sh