Shutdown but still see backgroud process

After bring the database down,
I issued ps -ef |grep smon...
I still see background processes.
To shutdown the database, I need to kill those processes manually.
Have you ever experience this ?

You see which processes? If they are oracleSID processes, then they are user processes and can be killed (they don't have to be, but they can be). If they are processes that belong to the database (SID_pmon, SID_smon, etc.) then they must be killed via sqlplus or svrmgrl with a shutdown command.

Unfortunately, they are not user processes.
I can not shutdown the database without killing smon, dbwr... processes manually.
See the following. Isn't it funny?

bash-2.03$ ps -ef |grep TEST
oracle 2086 1 0 17:06:43 ? 0:00 ora_lgwr_TEST
oracle 2092 1 0 17:06:43 ? 0:00 ora_reco_TEST
oracle 2082 1 0 17:06:43 ? 0:00 ora_dbw0_TEST
oracle 2090 1 0 17:06:43 ? 0:00 ora_smon_TEST
oracle 2080 1 0 17:06:42 ? 0:00 ora_pmon_TEST
oracle 2088 1 0 17:06:43 ? 0:00 ora_ckpt_TEST
oracle 2084 1 0 17:06:43 ? 0:00 ora_arch_TEST


SVRMGR> select * from v$database;
select * from v$database
*
ORA-01034: ORACLE not available


[Edited by chobo on 02-06-2001 at 05:42 PM]

Its not Clear. If you shutdown the database you can't see any of your background processes. If its not allowing you to shutdown the database, try using 'shutdown abort' which is equivalent of what you are thinking of to do (killing background processes). I can't predict what could be the result if you kill the background processes using 'kill' cmd.

Its clear with the latest post. Use 'ABORT'. seems the only option you have...

Shutdown abort gave me more funny result.

SVRMGRL> startup
ORACLE instance started.
Total System Global Area 17189296 bytes
Fixed Size 51632 bytes
Variable Size 10493952 bytes
Database Buffers 6553600 bytes
Redo Buffers 90112 bytes
Database mounted.
Database opened.

SVRMGRL> shutdown abort
ORACLE instance shut down.

ps -ef |grep smon
oracle 2247 1 0 17:41:09 ? 0:00 ora_smon_TEST
oracle 2090 1 0 17:06:43 ? 0:00 ora_smon_TEST

Now I see two smon processes.

Kill the left over smon processes using 'Kill' after bringing down the database using 'shutdown abort' which cleans up heldup background process.

It strange. I have never seen the situation like this.

I shutdown the DB and kill the all processes and then
restart the DB and shutdown the DB.
It seems OK right now.
It is strange.

shutdown abhort, open and shutdown immediate(cleanly this time). See if you still have any more processes left hanging.

Maybe your SMON is attempting to Coalence temp segments thats why it takes longer and the reason you get database isnt available is because the database isnt open. this happens if your SORT_AREA_SIZE is sized too small then SMON goes to disk and creates temporary segments which it has to release right before shuting down.

You can coalence your temp segments manually too.