Re: ORA-12805

On Apr 3, 6:39 am, "Michael42" <melliot..._at_yahoo.com> wrote:
> Hello,
>
> In an Oralce 10.2 R2 (10.2.03) database on a Solaris 10 system we are
> randomly getting ORA-12805 errors (ORA-12805: parallel query server
> died unexpectedly). We are not using RAC. Also, on both the
> tablespaces and effected tables the Parallel option is not enabled.
>
> Corresponding /udump trace files being generated that contain messages
> as shown below:
>
> SERVICE NAME: (SYS$USERS) 2007-04-02 12:04:34.330
> SESSION ID: (388.6351) 2007-04-02 12:04:34.330
> kxfpgidr
>
> Time out waiting for slave Join-Ack: passes:150 waits:150 csrret:3
> sleep
> time:200 timeouts:150 ghostposts 0
> start: 18960889658 end: 1896391150 diff:301492 ksl:301492 avg_ksl:2009
> old,new # of posts = 6,5
>
> *** 2007-04-02 12:04:34:331
> KSTDUMP: In-memory trace dump
> TIME(usecs):SEQ ...
>
> <<< 5 megabytes of memory dump follow>>>
>
> What can you recommend?
>
> Thanks,
>
> Michael

First I would get a service request into oracle on this.

Second I would take a look at all of the tables and indexes and view definitions involved. My shot in the dark is that parallel is specified somewhere in those ... Received on Tue Apr 03 2007 - 07:12:56 CDT