Skip to Main Content

Database Software

Announcement

For appeals, questions and feedback about Oracle Forums, please email oracle-forums-moderators_us@oracle.com. Technical questions should be asked in the appropriate category. Thank you!

Oracle Node Eviction in 2 node RAC on AIX platform

user11988068Jan 22 2011 — edited Jan 24 2011
All,

I have 2 node RAC on AIX 5.3 and Oracle 10G R2. I have CRS,ASM,DB seperate installations.
2 Nodes ==> Node1, Node2
My Node1 got evicted, I am trying to find the root cause. Open SR Sevirity-3 no response from Oracle.
what I checked.
1) ocssd.log and CRS alert log on serviving instance ( Because of our Log cleanup script bug , we lost the ocssd.log )
we fix the script and now we will not clean any CRS log except the logfile in client directory.
2) /etc/oracle/oprocd/*.log.* and lgl
not much information

I did not find the lgl<timestamp> logfiles. on evicted and serviving node.

lgl files are only available in only 11G? or We have missed any patch?? we do not have any patch applied on CRS home apart from 10.2.0.4.


3) errpt -a on the evicted node, which says user has ask for the shutdown, which SA says it was "oracle" user.
4) checked the alert and ocssd logs on evicted node.
5) checked for any interconnect bandwidth issues by using our openview system. not helpfull as it collectes the data every 5 minutes.
6) oprocd has 10000 ms time setting using the diagwait.
7) checked on the grid control no help di dnot see any major waits. except control file related, which are due to goldengate
8) around same time I see the serviving instance doing lot of buffer reading, Which I am co-relating to db_cache recovery from evicted node.

I want to know, is there any other thing I am missing to check?
How I can find the root cause about the node eviction?

any help is appriciated.
regards
bhushan
Comments
Locked Post
New comments cannot be posted to this locked post.
Post Details
Locked on Feb 21 2011
Added on Jan 22 2011
6 comments
1,235 views