-
Still learning the db, so this may seem like an elementary architecture question but here goes:
When I shutdown normal on the db bdwh I still have processes that are running. Why is it that all process are not shutdown when performing a graceful shutdown of the db? I am sure that most know this but the processes are as follows:
oracle 28478 1 0 14:27:42 ? 0:00 ora_smon_bdwh
oracle 28488 1 0 14:27:42 ? 0:00 ora_snp3_bdwh
dwload 1170 13509 0 14:36:11 pts/9 0:00 grep bdwh
oracle 28480 1 0 14:27:42 ? 0:00 ora_reco_bdwh
oracle 28476 1 0 14:27:41 ? 0:00 ora_ckpt_bdwh
oracle 28474 1 0 14:27:41 ? 0:00 ora_lgwr_bdwh
oracle 28486 1 0 14:27:42 ? 0:00 ora_snp2_bdwh
oracle 28472 1 0 14:27:41 ? 0:00 ora_dbw0_bdwh
oracle 28484 1 0 14:27:42 ? 0:00 ora_snp1_bdwh
oracle 28482 1 0 14:27:42 ? 0:00 ora_snp0_bdwh
oracle 28470 1 0 14:27:41 ? 0:00 ora_pmon_bdwh
-
If those processes are running, that means your database was not shutdown. If you shutdown normal, Oracle will wait for all existing connections to be disconnected. If you have persistent connections, the database never gets shut down. If you shutdown immediate, Oracle will gracefully terminate existing connections.
Jeff Hunter
-
Are there any risks to outright deleting these processes once the db is shutdown?
-
If those processes are running, that means your database was not shutdown.
Those processes will not be running if your database is down. You should not kill these procesess from the OS.
Jeff Hunter
-
Actually the database is shutdown. When I try and restart I get the following output:
Total System Global
Area 560785840 bytes
Fixed Size 51632 bytes
Variable Size 67100672 bytes
Database Buffers 491520000 bytes
Redo Buffers 2113536 bytes
ORA-01102: cannot mount database in EXCLUSIVE mode
SVRMGR> Disconnected.
SVRMGR> Server Manager complete.
------------
A DOC from metalink states that I must:
(1) shutdown the db
(2) kill all processes active for that SID
(3) Delete lk[SID] file in $ORACLE_HOME/dbs
(4) restart DB
What is the significants of the 'lk' file? Will this process endanger the database?
-
What is the database version and the platforum? First of all you should never delete the running processes, without the proper execution of shutdown. Is the database in archive log mode?
Sam
Thanx
Sam
Life is a journey, not a destination!
-
db version 8.0.6 on Solaris 7. Yes in archive mode but not a lot of activity on the db - mostly houses data. Maybe my limited experiences in confusing the question.
I shutdown the db, for this purpose, with a shutdown abort, startup restrict, shutdown normal. In some cases the db processes above still exist. What are the issues with killing the processes manually once the db is shutdown.
-
When you connect does it tell you that you are connected to an idle instance, if so you are shutdown.
Try 'select status from v$instance' to check the status.
when you get a message telling you 'cannot mount exclusive', it usually means that your database and instance are already up.
Shutdown normal will wait for all user session to disconnect before shutting down, is anyone else logged in ?
try 'shutdown immediate', this will shut down dataabase immediately, rolling back any uncommited transactions and kill user sessions.
If you kill any of he mandatory background processes while the database is open it will result in instance failure, the system will automatically recover itself on startup, this is not recommended.
Once you have eliminated all of the impossible,
whatever remains however improbable,
must be true.
-
Ok, I am sure that I am confused so let me re-phrase my situation and question:
(1) Shutdown db(bdwh) from server manager:
shutdown abort
startup restrict
shutdown immediate **
** did this using normal and immediate with the same results
Result from startup:
Total System Global
Area 560785840 bytes
Fixed Size 51632 bytes
Variable Size 67100672 bytes
Database Buffers 491520000 bytes
Redo Buffers 2113536 bytes
ORA-01102: cannot mount database in EXCLUSIVE mode
SVRMGR> Disconnected.
SVRMGR> Server Manager complete.
(2) From Solaris7 command line I grep for db process:
ps -ef |grep bdwh
Result from command:
oracle 28478 1 0 14:27:42 ? 0:00 ora_smon_bdwh
oracle 28488 1 0 14:27:42 ? 0:00 ora_snp3_bdwh
oracle 1170 13509 0 14:36:11 pts/9 0:00 grep bdwh
oracle 28480 1 0 14:27:42 ? 0:00 ora_reco_bdwh
oracle 28476 1 0 14:27:41 ? 0:00 ora_ckpt_bdwh
oracle 28474 1 0 14:27:41 ? 0:00 ora_lgwr_bdwh
oracle 28486 1 0 14:27:42 ? 0:00 ora_snp2_bdwh
oracle 28472 1 0 14:27:41 ? 0:00 ora_dbw0_bdwh
oracle 28484 1 0 14:27:42 ? 0:00 ora_snp1_bdwh
oracle 28482 1 0 14:27:42 ? 0:00 ora_snp0_bdwh
oracle 28470 1 0 14:27:41 ? 0:00 ora_pmon_bdwh
Now the database is shutdown.
The db is down and the processes above are still active. For the most part, I only have this problem when I try and run these server manager commands from a k-shell script. Why would it cause this problem? I have set up the environment and connect to internal fine. The simple test script follows:
#!/usr/bin/ksh
. ${HOME}/.profile
ORACLE_SID=${1}
export ORACLE_SID
ORACLE_HOME=/u01/app/oracle/product/8.0.6/
export ORACLE_HOME
# Shutdown and startup database
$ORACLE_HOME/bin/svrmgrl << EOF
connect internal
shutdown abort
startup restrict
shutdown immediate
startup
disconnect
exit
EOF
Any ideas?
-
Ah ha, now we get to the root problem.
Do these processes exist when you execute these commands from the command line instead of inside a script? I would bet not.
Jeff Hunter
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
|