Error 691 occuring on a 75GB ASE 12.0.0.1 database
Phil
requested to create a 75GB database on an ASE12 Intel
server.
O/S = Windows 2000 sp4, quad p3's, 3GB RAM
DB version = Adaptive Server Enterprise/12.0.0.1/P/SWR 9271
ESD 3/NT (IX86)/OS 4.0/1629/32bit/OPT/Wed Sep 06 14:53:20
2000
ASE installs as normal. I then created the database on 38
2G device fragments (disk is SAN based). Once the database
has been loaded with a copy of Prod (only 2GB data), the
Stress/Performance test begins. In the middle of testing,
database corruption occurs with error 691 being written to
the error log. Specifically, the error message reads as
such:
'Encountered invalid logical page '0' while accessing object
'14' in database '4'. This is an internal system error.
Please contact Sybase Technical Support.
where:
Object 14 = sysgams
database 4 = user database I created (not tempdb)
I am wondering if this is a bug. The database runs fine
prior to encountering this error. Once the error is
encountered, the database hangs. Upon bouncing the db
engine, the database is flagged as corrupt and needs to be
dropped using dbcc. I have searched the net and have found
suggestions of a 63GB db limit when error 691 indicates
sysgams is the object in question.
I am wondering if this is happening here? I am also
wondering if I need to apply a later EBF? Funny thing about
this is that the process that causes the error in the
Stress/Perf environment runs completely clean in Prod (Prod
db engine is exactly the same as what is in S/P).
Any help/suggestions would be greatly appreciated.
Phil
Mark A. Parsons
>
> 'Encountered invalid logical page '0' while accessing object
> '14' in database '4'. This is an internal system error.
> Please contact Sybase Technical Support.
>
> where:
> Object 14 = sysgams
> database 4 = user database I created (not tempdb)
>
> I am wondering if this is a bug. The database runs fine
> prior to encountering this error. Once the error is
> encountered, the database hangs. Upon bouncing the db
> engine, the database is flagged as corrupt and needs to be
> dropped using dbcc. I have searched the net and have found
> suggestions of a 63GB db limit when error 691 indicates
> sysgams is the object in question.
>
> I am wondering if this is happening here? I am also
> wondering if I need to apply a later EBF? Funny thing about
> this is that the process that causes the error in the
> Stress/Perf environment runs completely clean in Prod (Prod
> db engine is exactly the same as what is in S/P).
>
> Any help/suggestions would be greatly appreciated.
Have you checked the writeup for 691 in the Troubleshooting Guide?
In a nutshell ... the rec's are to
- check for other error messages in the log that might zero in on the
culprit
- check for hardware errors
- run a series of dbcc's to zero in on the problem
- contact Sybase TS
--
Mark A. Parsons
Iron Horse, Inc.
iron_...@NOSPAM.compuserve.com
mikhail
if you're getting the stacktrace, check if it matches the
one mentioned in the paper.
misha
mikhail
to solved cases, check case #10892388 - it looks like your
issue.
misha
p.s. if you cannot get there, i'll post the whole case.
> As part of a stress and performance exercise, I have been
> requested to create a 75GB database on an ASE12 Intel
> server
>
Phil
I have checked the writeup on error 691 in the
Troubleshooting Guide. Here are the responses to your
questions ...
- There are no other error messages in the log that point to
a potential culprit. Error 691 is written to the error log
with no corresponding stack trace as well.
- the o/s event viewer and other o/s logs do not indicate
any hardware issues
- all dbcc's recommended in the Troubleshooting Guide reveal
no corruption.
Looks like a call to Sybase TS is in order.
Thanks for your help.
Phil
I have access to solved cases. I just checked the case and
it looks like it might be worthwile to run the dbcc gam
command.
Thanks for your help.