Connections hang
Posted in 2005
Topics: Storage & Space Management, Server Administration, Logging & Checkpoints, Networking & sqlhosts Configuration, Versions, Editions & End-of-Life
Hi fellows,
from time to time, not quite often but believe me its pretty annoying, happens
that connections from remote servers or applications - every onsoctcp
connections - to our main IDS box just hang without any apparent reason.
Anyways local connections (e.g. a simple "dbaccess my_database") work fine!
What parameters onconfig should i check or tune? Any onstat option?
NETTYPE is set to soctcp,2,100,NET, all logical logs are backed up, there is
plenty of swap space (for the OS as well as for informix TMP dbspaces), CPU is
mostly idle, eliminated most busy sessions and some more at random pick from
"onstat -u" to reduce their number, chache ratios are OK, locks are at normal
level, network is working OK at full speed. Up to now, sincerely do not know
what else to check.
The box is IDS 9.40 on RedHat 7.3.
Any help is really welcome
Can you reproduce the problem? You could start by determining what is
happening to the session when the process is hung. That will tell you
whether it is trying to do something in the engine or point you in another
direction.
You cold also run the application with set explain on to determine whether
one sql is poor or whether you have too many sqls for your server to handle
in a reasonable time.
MW
> -----Original Message-----
> From: forum.subscriber@iiug.org
> [mailto:forum.subscriber@iiug.org] On Behalf Of ESTEBAN RIEZNIK
> Sent: Saturday, 16 April 2005 7:21 a.m.
> To: ids@iiug.org
> Subject: Connections hang [4728]
>
> Hi fellows,
>
> from time to time, not quite often but believe me its pretty
> annoying, happens that connections from remote servers or
> applications - every onsoctcp connections - to our main IDS
> box just hang without any apparent reason. Anyways local
> connections (e.g. a simple "dbaccess my_database") work fine!
>
> What parameters onconfig should i check or tune? Any onstat option?
>
> NETTYPE is set to soctcp,2,100,NET, all logical logs are
> backed up, there is plenty of swap space (for the OS as well
> as for informix TMP dbspaces), CPU is mostly idle, eliminated
> most busy sessions and some more at random pick from "onstat
> -u" to reduce their number, chache ratios are OK, locks are
> at normal level, network is working OK at full speed. Up to
> now, sincerely do not know what else to check.
>
> The box is IDS 9.40 on RedHat 7.3.
>
> Any help is really welcome
>
>
>
>
You might have tried all the below, if not
1. What is the status of the engine "onstat - ". Is it online or is it
in Blocked CKPT state. Might be it is waiting on a check point.
2. What is the status of each session "onstat -u ", third column flags.
The first character/flag would tell you the status of the session. You
can get the explanation for each status from the manuals. It could be
C/Y/-.
3. Have you tried onstat -g ntu. It tells you about the threads.
4. onstat -g act. This tells you whether a thread is running. If you
don't see them moving, then that session/thread could be loading your
CPU.
5. onstat -g wai. This tells you the threads that are waiting.
6. When it hangs, do you recycle the instance. Try getting some
statistics. "Onstat -a" could take some time and space.
Thank you,
Kannan Thirugnanam
-----Original Message-----
From: forum.subscriber@iiug.org [mailto:forum.subscriber@iiug.org] On
Behalf Of Murray Wood....
Sent: Sunday, April 17, 2005 6:21 PM
To: ids@iiug.org
Subject: RE: Connections hang [4730]
Can you reproduce the problem? You could start by determining what is
happening to the session when the process is hung. That will tell you
whether it is trying to do something in the engine or point you in
another
direction.
You cold also run the application with set explain on to determine
whether
one sql is poor or whether you have too many sqls for your server to
handle
in a reasonable time.
MW
> -----Original Message-----
> From: forum.subscriber@iiug.org
> [mailto:forum.subscriber@iiug.org] On Behalf Of ESTEBAN RIEZNIK
> Sent: Saturday, 16 April 2005 7:21 a.m.
> To: ids@iiug.org
> Subject: Connections hang [4728]
>
> Hi fellows,
>
> from time to time, not quite often but believe me its pretty
> annoying, happens that connections from remote servers or
> applications - every onsoctcp connections - to our main IDS
> box just hang without any apparent reason. Anyways local
> connections (e.g. a simple "dbaccess my_database") work fine!
>
> What parameters onconfig should i check or tune? Any onstat option?
>
> NETTYPE is set to soctcp,2,100,NET, all logical logs are
> backed up, there is plenty of swap space (for the OS as well
> as for informix TMP dbspaces), CPU is mostly idle, eliminated
> most busy sessions and some more at random pick from "onstat
> -u" to reduce their number, chache ratios are OK, locks are
> at normal level, network is working OK at full speed. Up to
> now, sincerely do not know what else to check.
>
> The box is IDS 9.40 on RedHat 7.3.
>
> Any help is really welcome
>
>
>
>
The most
suspicious was the following, several annoyng 27001 error:
19:40:57 Maximum server connections 28
19:41:15 listener-thread: err = -27001: oserr = 0: errstr = : Read
error occurred during connection attempt.
19:42:18 listener-thread: err = -27001: oserr = 0: errstr = : Read
error occurred during connection attempt.
The problem stayed even after tuning poll threads stuff, and even after
rebooting server, with just very few connections. onstat, netstat et al.
showed nothing suspicious. The listener thread was still screwed up.
Due to lack of time for further in-depth analysis, a mate changed the
port numbers of all the instances and IDS surprisingly started working
fine again. We still can not figure out what could have happened. Never
before in more than 1 yr.
We *** suspect *** of a crippled remote server / router on the same day
possibly sending corrupted data, but not sure yet.
Unfortunately we can not reproduce the problem, but it would help for
sure. Thanx a lot you guys for your help. I'll have your advices in
account for next time.
All the best.
-----Mensaje original-----
De: forum.subscriber@iiug.org [mailto:forum.subscriber@iiug.org] En
nombre de Thirugnanam....
Enviado el: Lunes, 18 de Abril de 2005 11:13 a.m.
Para: ids@iiug.org
Asunto: RE: Connections hang [4735]
You might have tried all the below, if not
1. What is the status of the engine "onstat - ". Is it online or is it
in Blocked CKPT state. Might be it is waiting on a check point.
2. What is the status of each session "onstat -u ", third column flags.
The first character/flag would tell you the status of the session. You
can get the explanation for each status from the manuals. It could be
C/Y/-.
3. Have you tried onstat -g ntu. It tells you about the threads.
4. onstat -g act. This tells you whether a thread is running. If you
don't see them moving, then that session/thread could be loading your
CPU.
5. onstat -g wai. This tells you the threads that are waiting.
6. When it hangs, do you recycle the instance. Try getting some
statistics. "Onstat -a" could take some time and space.
Thank you,
Kannan Thirugnanam
-----Original Message-----
From: forum.subscriber@iiug.org [mailto:forum.subscriber@iiug.org] On
Behalf Of Murray Wood....
Sent: Sunday, April 17, 2005 6:21 PM
To: ids@iiug.org
Subject: RE: Connections hang [4730]
Can you reproduce the problem? You could start by determining what is
happening to the session when the process is hung. That will tell you
whether it is trying to do something in the engine or point you in
another
direction.
You cold also run the application with set explain on to determine
whether
one sql is poor or whether you have too many sqls for your server to
handle
in a reasonable time.
MW
> -----Original Message-----
> From: forum.subscriber@iiug.org
> [mailto:forum.subscriber@iiug.org] On Behalf Of ESTEBAN RIEZNIK
> Sent: Saturday, 16 April 2005 7:21 a.m.
> To: ids@iiug.org
> Subject: Connections hang [4728]
>
> Hi fellows,
>
> from time to time, not quite often but believe me its pretty
> annoying, happens that connections from remote servers or
> applications - every onsoctcp connections - to our main IDS
> box just hang without any apparent reason. Anyways local
> connections (e.g. a simple "dbaccess my_database") work fine!
>
> What parameters onconfig should i check or tune? Any onstat option?
>
> NETTYPE is set to soctcp,2,100,NET, all logical logs are
> backed up, there is plenty of swap space (for the OS as well
> as for informix TMP dbspaces), CPU is mostly idle, eliminated
> most busy sessions and some more at random pick from "onstat
> -u" to reduce their number, chache ratios are OK, locks are
> at normal level, network is working OK at full speed. Up to
> now, sincerely do not know what else to check.
>
> The box is IDS 9.40 on RedHat 7.3.
>
> Any help is really welcome
>
>
>
>
AVISO LEGAL:
Esta informacion es privada y confidencial y esta dirigida unicamente a su
destinatario. Si usted no es el destinatario original de este mensaje y por
este medio pudo acceder a dicha informacion por favor elimine el mensaje. La
distribucion o copia de este mensaje esta estrictamente prohibida. Esta
comunicacion es solo para propositos de informacion y no debe ser considerada
como propuesta, aceptacion ni como una declaracion de voluntad oficial de
REPSOL YPF S.A. y/o subsidiarias y/o afiliadas. La transmision de e-mails no
garantiza que el correo electronico sea seguro o libre de error. Por
consiguiente, no manifestamos que esta informacion sea completa o precisa.
Toda informacion esta sujeta a alterarse sin previo aviso.
This information is private and confidential and intended for the recipient
only. If you are not the intended recipient of this message you are hereby
notified that any review, dissemination, distribution or copying of this
message is strictly prohibited. This communication is for information purposes
only and shall not be regarded neither as a proposal, acceptance nor as a
statement of will or official statement from REPSOL YPF S.A. and/or
subsidiaries and/or affiliates. Email transmission cannot be guaranteed to be
secure or error-free. Therefore, we do not represent that this information is
complete or accurate and it should not be relied upon as such. All information
is subject to change without notice.
When the
conmnection hang, are all clients are affecte dat the same time ?
Or is the problem punctual , ie, one client is affetcd now, then a
different client
the next time and so on?
What is the status of the corresponding Informix session when it get hung
- onstat -g ses <sid>
(to get the <sid> type onstat -g ses (without an additional argument an
search for
the hung process' PID)
What is verison your client connect libraries?
Regards
Tilman
--
Tilman Model-Bosch
IBM Data Managment Solutions, Informix Advanced Support
c\\\\o SAP AG
TECHDEV 05
Neurrotstr.16
69190 Walldorf
forum.subscriber@iiug.org wrote on 18/04/2005 16:12:34:
> You might have tried all the below, if not
> 1. What is the status of the engine "onstat - ". Is it online or is it
> in Blocked CKPT state. Might be it is waiting on a check point.
> 2. What is the status of each session "onstat -u ", third column flags.
> The first character/flag would tell you the status of the session. You
> can get the explanation for each status from the manuals. It could be
> C/Y/-.
> 3. Have you tried onstat -g ntu. It tells you about the threads.
> 4. onstat -g act. This tells you whether a thread is running. If you
> don't see them moving, then that session/thread could be loading your
> CPU.
> 5. onstat -g wai. This tells you the threads that are waiting.
> 6. When it hangs, do you recycle the instance. Try getting some
> statistics. "Onstat -a" could take some time and space.
>
>
> Thank you,
> Kannan Thirugnanam
>
> -----Original Message-----
> From: forum.subscriber@iiug.org [mailto:forum.subscriber@iiug.org] On
> Behalf Of Murray Wood....
> Sent: Sunday, April 17, 2005 6:21 PM
> To: ids@iiug.org
> Subject: RE: Connections hang [4730]
>
> Can you reproduce the problem? You could start by determining what is
> happening to the session when the process is hung. That will tell you
> whether it is trying to do something in the engine or point you in
> another
> direction.
> You cold also run the application with set explain on to determine
> whether
> one sql is poor or whether you have too many sqls for your server to
> handle
> in a reasonable time.
>
> MW
>
> > -----Original Message-----
> > From: forum.subscriber@iiug.org
> > [mailto:forum.subscriber@iiug.org] On Behalf Of ESTEBAN RIEZNIK
> > Sent: Saturday, 16 April 2005 7:21 a.m.
> > To: ids@iiug.org
> > Subject: Connections hang [4728]
> >
> > Hi fellows,
> >
> > from time to time, not quite often but believe me its pretty
> > annoying, happens that connections from remote servers or
> > applications - every onsoctcp connections - to our main IDS
> > box just hang without any apparent reason. Anyways local
> > connections (e.g. a simple "dbaccess my_database") work fine!
> >
> > What parameters onconfig should i check or tune? Any onstat option?
> >
> > NETTYPE is set to soctcp,2,100,NET, all logical logs are
> > backed up, there is plenty of swap space (for the OS as well
> > as for informix TMP dbspaces), CPU is mostly idle, eliminated
> > most busy sessions and some more at random pick from "onstat
> > -u" to reduce their number, chache ratios are OK, locks are
> > at normal level, network is working OK at full speed. Up to
> > now, sincerely do not know what else to check.
> >
> > The box is IDS 9.40 on RedHat 7.3.
> >
> > Any help is really welcome
> >
> >
> >
> >
>
>
>
>
>