-
Hi there
I had a slight problem with my backup on my prod. dbase last night which ended when the database would not shut down before doing the cold backup. Looking thru the trace file revealed the line:
SHUTDOWN: waiting for active calls to complete.
I have checked through Metalink and this of course comes down to Oracle awaiting the clean-up of uncommitted transactions and/or the clean-up of existing processes. This is the first time this problem has happened in 18 months. Have others had this problem and what did they do to ensure it didnt happen. Thanks in advance
Bruce
-
Are you using shutdown immediate or shutdown normal?
Jeff Hunter
-
its shutdown immediate. Sorry for not making it clearer. Any ideas???
-
you will have to write a logic within your shutdown script to sleep your script for 15 mins after shutdown immediate...then check for pmon to see if its up and grep the alert_log for Shutdown: waiting for calls to complete. If true then shutdown abort...and wait about 2 mins then startup the database and email and page you.
thats what i did.
let me knwo if you find a workaround
"High Salaries = Happiness = Project Success."
-
Hi there
Thanks for the reply. As I inherited this backup script which is quite complex I would need some help to sort it out. I like the idea which you have suggested. Metalink were not that helpful
Before issuing the shutdown command issue:
Alter session set events '10400 trace name context forever, level 1';
shutdown immediate
This will produce a few trace files in your "user_dump_dest" directory."user_dump_dest" is an "init.ora" parameter. The trace file will be a systemstate dump. Upload the trace files. From the trace files, it can be seen that what processes are still active.
Many Thanks!
-
you can setup that
"Alter session set events '10400 trace name context forever, level 1'; shutdown immediate " in your shutdown script.
"High Salaries = Happiness = Project Success."
-
I've checked the alter session synthax out using Oracle8 the complete ref, but
there's no reference to "event", could someone explain this synthax?
Thanks.
-
Hi there
I hope I have not confused anyone. The official line from Metalink is:
This message is written to the alert log when the database is waiting for pmon to clean up processes, but pmon is unable to
clean them. The client connections to the server are causing the shutdown immediate or normal to hang. The workaround in such a situation will be to kill such processes ( Refer Note:1039146.6).
If there were no client connection at the time of issuing this command, then please do the following.
Before issuing the shutdown command issue:
Alter session set events '10400 trace name context forever, level 1';
shutdown immediate
This will produce a few trace files in your "user_dump_dest" directory."user_dump_dest" is an "init.ora" parameter. The trace file will be a systemstate dump. Upload the trace files. From the trace files, it can be seen that what processes are still active.
The other workaround in this case will be to shutdown with:
shutdown abort
startup restrict
shutdown normal
Since it happened only once, I would suspect that there might be some long running transaction going on in any of the sessions that were present at the time of shutdown. This could have caused because the shutdown was waiting for this transaction to roll back completely. There are no known bugs on this. The other possibility of this could be that there are some blocking processes present in the system, which caused some of the other sessions to hang. The trace file generated by event 10400 will help to analyze such a scenario
I hope this info helps. Thanks for your input.
Posting Permissions
- You may not post new threads
- You may not post replies
- You may not post attachments
- You may not edit your posts
-
Forum Rules
|
Click Here to Expand Forum to Full Width
|