Re: More baby steps into Ora on unix - what is this?

They're the processes associated with the Oracle connections from "remote" clients. If you've shut the databases down cleanly, then they should have been cleaned up. The fact that they aren't may be down to the way the system was shutdown, or "bad" client disconnections. Or, it may be a bug. Patching to 8.1.7.4 _might_ fix it.

If the database is definitely shutting down cleanly, then there should be no harm in killing these processes off.

 ps -ef | grep "oracleCMDM1240" | awk '{ print $2 }' | xargs kill -9

should do it quickly and easily. But make sure the awk part outputs the correct process ID's first: you don't want to kill the wrong process.

Ed Stevens <nospam_at_noway.nohow> wrote in news:la2h7v4b4v4pfk3vajvri6tfj3v8ajrkd2_at_4ax.com:

> Platform; Oracle 8.1.7.0 EE on Solaris
>
> After shutting down the db's and stopping the listener,
>
> ps -ef|grep "ora" | grep -v grep
>
> yeilds several pages of the following:
>
> oracle 10676 1 0 07:42:12 ? 0:00 oracleCMDM1240 (LOCAL=NO)
> oracle 10721 1 0 07:42:13 ? 0:00 oracleCMDM1240 (LOCAL=NO)
> oracle 10727 1 0 07:42:14 ? 0:00 oracleCMDM1240 (LOCAL=NO)
> oracle 10670 1 0 07:42:12 ? 0:00 oracleCMDM1240 (LOCAL=NO)
>
> the CMDM1240 is the name of one of two SIDs on this box. It is the
> only one referenced in the several pages of output.
Received on Wed Mar 19 2003 - 11:05:30 CST