BUG #9604: Unable to access table remotely

On 03/17/2014 06:38 PM, Michael.buccetti@airgas.com wrote:

The following bug has been logged on the website:

Bug reference: 9604
Logged by: Michael Buccetti
Email address: Michael.buccetti@airgas.com
PostgreSQL version: 9.1.0

Next week you need to upgrade to 9.1.13. You're missing two and a half
years of bug fixes.

Operating system: OpenSuse
Description:

We have a java web app running on Tomcat 7 that accesses data on a Postgres
database. For some reason, we get the following error unexpectedly, causing
a crucial database query to fail:

ERROR: could not open file "base/16397/106156081": No such file or
directory.

We installed PGAdmin on the machine and verified that the error shows up
when querying a specific table.

We are able to run the same query on various other machines without an
issue, and sometimes it looks like the issue clears itself up, but we have
not been able to pinpoint the fix to anything specific that we tried.

Have you seen any issues like this? We thought that maybe the installation
of Tomcat might have corrupted something on the system that affected both
PGAdmin and the jdbc access to the database, but uninstalling Tomcat did not
fix the issue.

Is there any chance that that is an unlogged table and the error occurs
after a crash?

I've experienced on two separate sites an error like this on unlogged
tables after a crash, but I can't reproduce it and reading the code
seems straightforward.

--
Vik

--
Sent via pgsql-bugs mailing list (pgsql-bugs@postgresql.org)
To make changes to your subscription:
http://www.postgresql.org/mailpref/pgsql-bugs

On Tue, Mar 18, 2014 at 4:38 AM, <Michael.buccetti@airgas.com> wrote:

The following bug has been logged on the website:

Bug reference: 9604
Logged by: Michael Buccetti
Email address: Michael.buccetti@airgas.com
PostgreSQL version: 9.1.0
Operating system: OpenSuse
Description:

We have a java web app running on Tomcat 7 that accesses data on a Postgres
database. For some reason, we get the following error unexpectedly,
causing
a crucial database query to fail:

ERROR: could not open file "base/16397/106156081": No such file or
directory.

Could you please help us know if the above file is missing from the
location ( data-directory/base/16397 ) on the host where the database is
hosted ?

You might be knowing this : The file "106156081" belongs to a particular
table in the database with OID 16397.

If you have audit logs - did you notice if the file was accidentally
deleted or may be file system permissions/ownership has changed ?

Or any activity has happened which would not let the database user access
the file from the OS level ?

Regards,
Venkata Balaji N

Sr. Database Administrator
Fujitsu Australia

Hello,

Thank you for getting back to me so quickly. One of our network gurus started digging into the issue last night and found that there was a problem with our Pacemaker syncing. Essentially it failed to sync properly and promoted a slave node to Master and gave two different nodes the “master” IP address. Also, the table we were accessing was an unlogged table. We believe these issues caused our problem. Once the syncing issue was resolved and the IP addresses were corrected, the issue cleared itself up.

Going forward we are going to monitor the situation and see if the syncing issue repeats.

Thank you again,

Michael Buccetti

On Mar 17, 2014, at 7:21 PM, Venkata Balaji Nagothi <vbnpgc@gmail.com<mailto:vbnpgc@gmail.com>> wrote:

On Tue, Mar 18, 2014 at 4:38 AM, <Michael.buccetti@airgas.com<mailto:Michael.buccetti@airgas.com>> wrote:
The following bug has been logged on the website:

Bug reference: 9604
Logged by: Michael Buccetti
Email address: Michael.buccetti@airgas.com<mailto:Michael.buccetti@airgas.com>
PostgreSQL version: 9.1.0
Operating system: OpenSuse
Description:

We have a java web app running on Tomcat 7 that accesses data on a Postgres
database. For some reason, we get the following error unexpectedly, causing
a crucial database query to fail:

ERROR: could not open file "base/16397/106156081": No such file or
directory.

Could you please help us know if the above file is missing from the location ( data-directory/base/16397 ) on the host where the database is hosted ?

You might be knowing this : The file "106156081" belongs to a particular table in the database with OID 16397.

If you have audit logs - did you notice if the file was accidentally deleted or may be file system permissions/ownership has changed ?

Or any activity has happened which would not let the database user access the file from the OS level ?

Regards,
Venkata Balaji N

Sr. Database Administrator
Fujitsu Australia