Oracle FAQ Your Portal to the Oracle Knowledge Grid
HOME | ASK QUESTION | ADD INFO | SEARCH | E-MAIL US
 

Home -> Community -> Usenet -> c.d.o.server -> Re: Q: complete and utter weird database errors (crashing!)

Re: Q: complete and utter weird database errors (crashing!)

From: Sylvie Bérubé <sberube_at_cam.org>
Date: 1997/01/12
Message-ID: <32D961BA.2FEE@cam.org>#1/1

Gary Assa wrote:
>

<snip>

> How can I find out what lies at that block, because almost every time that
> that database crashes, we get this block showing up. We have even moved
> tablespace around to different disks, and still the same block appeared.

<snip>
> .
> .
> .
> Thu Jan 2 23:01:25 1997
> Thread 1 advanced to log sequence 455
> Current log# 2 seq# 455 mem# 0: /oracle7/data/TX01/dbf/redo_02.dbf
> Current log# 2 seq# 455 mem# 1: /raid0/TX01/dbf/redo_02b.dbf
> Thu Jan 2 23:20:10 1997
>
> ******* here is that KCF error
>
> KCF: write/open error dba=80000aa block=aa online=1
> file=2 /raid1/TX01/dbf/rbs_01.dbf
> error=7376 txt: 'SVR4 Error: 5: I/O error
> Additional information: 170
> Additional information: 3'
> Automatic datafile offline due to write error on
> file 2: /raid1/TX01/dbf/rbs_01.dbf
> KCF: write/open error dba=180003fd block=3fd online=1
> file=6 /raid1/TX01/dbf/ts02_01.dbf
> error=7376 txt: 'SVR4 Error: 5: I/O error
> Additional information: 1021
> Additional information: 3'
> Automatic datafile offline due to write error on
> file 6: /raid1/TX01/dbf/ts02_01.dbf
> Thu Jan 2 23:22:01 1997
> Errors in file /oracle7/admin/TX01/bdump/tx01_smon_726.trc:
> ORA-00376: file 2 cannot be read at this time
> ORA-01110: data file 2: '/raid1/TX01/dbf/rbs_01.dbf'
> .

<snip>

Gary,

One thing could be that your rollback tablespace file is corrupted at this block. As it seeems to be the rollback tablespace (probably without data :-) ) ... did you tried putting rbs offline, then dropping the offline tablespace to recreate it completely with new rollback segments? Test first that your pack (hard disk is not damaged) before recreating the file but it shouldn't be as you moved the file around to another pack without getting rid of the problem.

I am thinking that maybe as one of your rollback segments in that tablespace increase in size it needs to write to that block and since the file is corrupted, you get that error message. Having a file corrupted Oracle puts it offline immediately which is the reason of the second part of your message (you already probably know that, sorry!)

Sylvie Bérubé
sberube_at_aircanada.ca Received on Sun Jan 12 1997 - 00:00:00 CST

Original text of this message

HOME | ASK QUESTION | ADD INFO | SEARCH | E-MAIL US