Hanging checkpoint
A DBA on IDS 9.30.FC2 under Solaris 8 reported a checkpoint that hung, blocking logins and queries until the server was bounced; logical logs weren't full and the online.log simply stopped after a normal fuzzy checkpoint entry. Respondents asked for version/cause details and advised that, if it recurs, he capture diagnostics before shutting down -- onstat -a, onstat -g all, onstat -g stk for each session, plus a shared-memory dump -- and open a case with Informix Support. No root cause or fix is recorded in the thread.
Auto-generated by DrWatson from the posts below — may be imperfect; read the full thread.
Recently we had an issue where a checkpoint was hanging and preventing users from logging on and running any type of query on the machine. I had to shutdown the Informix server and start it back up again to get it to a useable state. Does anyone have any ideas on how to prevent something like this in the future? ______________________ Keith Schleicher DBA - ACN Tech Services Phone: 920-405-7955 NETS: 8-270-7955 email: kschleicher@acnielsen.com
Version and platform information will help. Also any information you have about what caused the checkpoint to hang.... Were the logical logs full, a client task in a critical section had been paused (SIGSTOP), etc. Art S. Kagel ----- Original Message ----- From: .... Schleicher <KSchleicher@acnielsen.com> At: 6/24 10:42 > Recently we had an issue where a checkpoint was hanging and preventing users > from logging on and running any type of query on the machine. I had to > shutdown the Informix server and start it back up again to get it to a > useable state. Does anyone have any ideas on how to prevent something like > this in the future? > ______________________ > Keith Schleicher > DBA - ACN Tech Services > Phone: 920-405-7955 > NETS: 8-270-7955 > email: kschleicher@acnielsen.com
Running SunOS 5.8 IDS Version 9.30.FC2 Logical logs weren't full, not even close. No users were doing any critical task. Nothing in the log indicated any issues other than it just seemed to stop suddenly. Suddenly the log stopped recording any checkpoints. ______________________ Keith Schleicher DBA - ACN Tech Services Phone: 920-405-7955 NETS: 8-270-7955 email: kschleicher@acnielsen.com -----Original Message----- From: ART KAGEL, BLOOMBERG/ 499 PARK [mailto:KAGEL@bloomberg.net] Sent: Thursday, June 24, 2004 10:01 AM To: Schleicher, Keith M. Subject: Re: Hanging checkpoint [3148] Version and platform information will help. Also any information you have about what caused the checkpoint to hang.... Were the logical logs full, a client task in a critical section had been paused (SIGSTOP), etc. Art S. Kagel ----- Original Message ----- From: .... Schleicher <KSchleicher@acnielsen.com> At: 6/24 10:42 > Recently we had an issue where a checkpoint was hanging and preventing users > from logging on and running any type of query on the machine. I had to > shutdown the Informix server and start it back up again to get it to a > useable state. Does anyone have any ideas on how to prevent something like > this in the future? > ______________________ > Keith Schleicher > DBA - ACN Tech Services > Phone: 920-405-7955 > NETS: 8-270-7955 > email: kschleicher@acnielsen.com
In
a time long ago, when I was working at Informix, they used to tell us to
advise people when something like this happened to do the following:
1. onstat -a
2. onstat -g all
3. onstat -g stk for every session
If this happens again, generate this information and open a case with Informix
Support.
Take care.
Clifton
PS. They may want more or less information now ... who knows :) CB
"Schleicher,...." <KSchleicher@ACNielsen.com> wrote:
Running SunOS 5.8
IDS Version 9.30.FC2
Logical logs weren't full, not even close.
No users were doing any critical task.
Nothing in the log indicated any issues other than it just seemed to stop
suddenly. Suddenly the log stopped recording any checkpoints.
______________________
Keith Schleicher
DBA - ACN Tech Services
Phone: 920-405-7955
NETS: 8-270-7955
email: kschleicher@acnielsen.com
-----Original Message-----
From: ART KAGEL, BLOOMBERG/ 499 PARK [mailto:KAGEL@bloomberg.net]
Sent: Thursday, June 24, 2004 10:01 AM
To: Schleicher, Keith M.
Subject: Re: Hanging checkpoint [3148]
Version and platform information will help. Also any information you have
about
what caused the checkpoint to hang.... Were the logical logs full, a client
task in a critical section had been paused (SIGSTOP), etc.
Art S. Kagel
----- Original Message -----
From: .... Schleicher
At: 6/24 10:42
> Recently we had an issue where a checkpoint was hanging and preventing
users
> from logging on and running any type of query on the machine. I had to
> shutdown the Informix server and start it back up again to get it to a
> useable state. Does anyone have any ideas on how to prevent something
like
> this in the future?
> ______________________
> Keith Schleicher
> DBA - ACN Tech Services
> Phone: 920-405-7955
> NETS: 8-270-7955
> email: kschleicher@acnielsen.com
Hi, can you send us the last lines of message log? Because I had some hangs without evidence and message log last lines was "14:53:00 Checkpoint loguniq 372, logpos 0xe727f4 " Mensaje citado por "Schleicher,...." <KSchleicher@ACNielsen.com>: > Running SunOS 5.8 > IDS Version 9.30.FC2 > Logical logs weren't full, not even close. > No users were doing any critical task. > Nothing in the log indicated any issues other than it just seemed to stop > suddenly. Suddenly the log stopped recording any checkpoints. > > ______________________ > Keith Schleicher > DBA - ACN Tech Services > Phone: 920-405-7955 > NETS: 8-270-7955 > email: kschleicher@acnielsen.com > > > -----Original Message----- > From: ART KAGEL, BLOOMBERG/ 499 PARK [mailto:KAGEL@bloomberg.net] > Sent: Thursday, June 24, 2004 10:01 AM > To: Schleicher, Keith M. > Subject: Re: Hanging checkpoint [3148] > > > Version and platform information will help. Also any information you have > about > what caused the checkpoint to hang.... Were the logical logs full, a > client > > task in a critical section had been paused (SIGSTOP), etc. > > Art S. Kagel > ----- Original Message ----- > From: .... Schleicher <KSchleicher@acnielsen.com> > At: 6/24 10:42 > > > Recently we had an issue where a checkpoint was hanging and preventing > users > > from logging on and running any type of query on the machine. I had to > > shutdown the Informix server and start it back up again to get it to a > > useable state. Does anyone have any ideas on how to prevent something > like > > this in the future? > > ______________________ > > Keith Schleicher > > DBA - ACN Tech Services > > Phone: 920-405-7955 > > NETS: 8-270-7955 > > email: kschleicher@acnielsen.com > > > > > > >
05:05:55 Fuzzy Checkpoint Completed: duration was 1 seconds, 7038 buffers not flushed. 05:05:55 Checkpoint loguniq 1134, logpos 0x6c1c124 05:05:55 Maximum server connections 13 ---------------------------------------------------- These are the last lines in the online.log that were written before the checkpoint was causing blocks. ______________________ Keith Schleicher DBA - ACN Tech Services Phone: 920-405-7955 NETS: 8-270-7955 email: kschleicher@acnielsen.com -----Original Message----- From: pamadeo@cespi.unlp.edu.ar [mailto:pamadeo@cespi.unlp.edu.ar] Sent: Thursday, June 24, 2004 1:11 PM To: Schleicher, Keith M. Cc: ids@iiug.org Subject: RE: Hanging checkpoint [3150] Hi, can you send us the last lines of message log? Because I had some hangs without evidence and message log last lines was "14:53:00 Checkpoint loguniq 372, logpos 0xe727f4 " Mensaje citado por "Schleicher,...." <KSchleicher@ACNielsen.com>: > Running SunOS 5.8 > IDS Version 9.30.FC2 > Logical logs weren't full, not even close. > No users were doing any critical task. > Nothing in the log indicated any issues other than it just seemed to stop > suddenly. Suddenly the log stopped recording any checkpoints. > > ______________________ > Keith Schleicher > DBA - ACN Tech Services > Phone: 920-405-7955 > NETS: 8-270-7955 > email: kschleicher@acnielsen.com > > > -----Original Message----- > From: ART KAGEL, BLOOMBERG/ 499 PARK [mailto:KAGEL@bloomberg.net] > Sent: Thursday, June 24, 2004 10:01 AM > To: Schleicher, Keith M. > Subject: Re: Hanging checkpoint [3148] > > > Version and platform information will help. Also any information you have > about > what caused the checkpoint to hang.... Were the logical logs full, a > client > > task in a critical section had been paused (SIGSTOP), etc. > > Art S. Kagel > ----- Original Message ----- > From: .... Schleicher <KSchleicher@acnielsen.com> > At: 6/24 10:42 > > > Recently we had an issue where a checkpoint was hanging and preventing > users > > from logging on and running any type of query on the machine. I had to > > shutdown the Informix server and start it back up again to get it to a > > useable state. Does anyone have any ideas on how to prevent something > like > > this in the future? > > ______________________ > > Keith Schleicher > > DBA - ACN Tech Services > > Phone: 920-405-7955 > > NETS: 8-270-7955 > > email: kschleicher@acnielsen.com > > > > > > >
even if
you have time before taking the engine down get a shared memory
dump, that would help.
esteban.-
At 01:43 p.m. 24/06/2004 -0400, Clifton M. Bean wrote:
>In a time long ago, when I was working at Informix, they used to tell us
>to advise people when something like this happened to do the following:
>
>1. onstat -a
>2. onstat -g all
>3. onstat -g stk for every session
>
>If this happens again, generate this information and open a case with
>Informix Support.
>
>Take care.
>Clifton
>
>PS. They may want more or less information now ... who knows :) CB
>
>"Schleicher,...." <KSchleicher@ACNielsen.com> wrote:
>Running SunOS 5.8
>IDS Version 9.30.FC2
>Logical logs weren't full, not even close.
>No users were doing any critical task.
>Nothing in the log indicated any issues other than it just seemed to stop
>suddenly. Suddenly the log stopped recording any checkpoints.
>
>______________________
>Keith Schleicher
>DBA - ACN Tech Services
>Phone: 920-405-7955
>NETS: 8-270-7955
>email: kschleicher@acnielsen.com
>
>
>-----Original Message-----
>From: ART KAGEL, BLOOMBERG/ 499 PARK [mailto:KAGEL@bloomberg.net]
>Sent: Thursday, June 24, 2004 10:01 AM
>To: Schleicher, Keith M.
>Subject: Re: Hanging checkpoint [3148]
>
>
>Version and platform information will help. Also any information you have
>about
>what caused the checkpoint to hang.... Were the logical logs full, a client
>
>task in a critical section had been paused (SIGSTOP), etc.
>
>Art S. Kagel
>----- Original Message -----
>From: .... Schleicher
>At: 6/24 10:42
>
> > Recently we had an issue where a checkpoint was hanging and preventing
>users
> > from logging on and running any type of query on the machine. I had to
> > shutdown the Informix server and start it back up again to get it to a
> > useable state. Does anyone have any ideas on how to prevent something
>like
> > this in the future?
> > ______________________
> > Keith Schleicher
> > DBA - ACN Tech Services
> > Phone: 920-405-7955
> > NETS: 8-270-7955
> > email: kschleicher@acnielsen.com