Thread: Procedure for fixing and repair multiple instances

Procedure for fixing and repair multiple instances

From
Luis Marin
Date:
Hello,

Please refresh my memory,

I have four instances, configured :

Server AAAA   /db1      instance 1    instance 2
                          /db2      instance 3    instance 4

Postgresql 9.3 running on Debian 7,

My disk array failed and recovered itself, all disk partitions mounted ok, but i have to run fsck in /db1 and /db2,
 I could manually start instance 1 whitout message errors, but instance 2 announces:

       /db1/postgresq9.3/cal/ is not a valid data directory
      file /db1/postgresql9.3/cal/PG_VERSION does not exist

So Instance 1 started ok but Instance 2 does not start, how I fix this error ? I try to re-create this file manually  but later mention to PG_CONTROL does not exist started to appear, this give me the idea that some kind of corruption damaged this directory.

Am I comdenmed to re-install again ???
What could be the best method without touching the instance that is working ok ?

Instance 3 and Instance 4 started ok but several messages like this:
   
        Statistics file pg_stat_tmp/db_12034.stat is corrupted

in their respective log files,

During the fsck processs for /db2 I could see several messages like, specially during phase 2 check directory structure:

ENTRY  'dbo.stat'  which is in /postgresql/9.3/qa/pg_stat_tmp has a incorrect file type (it was 1 but should be 2) Fix(y) ?

if this is related, how I could clean this messages or fix it ?

Thanks a lot

Any idea or recommendation is welcome




Re: Procedure for fixing and repair multiple instances

From
Scott Whitney
Date:

You might try select pg_stat_reset() on the instances complaining about the bad stat files.


I'm a little unclear on what you're calling an instance, here. Can you clarify?


What is db1, what is db2? What are instances 1-4?


From: pgsql-admin-owner@postgresql.org <pgsql-admin-owner@postgresql.org> on behalf of Luis Marin <luismarinaray@gmail.com>
Sent: Thursday, September 24, 2015 11:30 AM
To: pgsql-admin@postgresql.org
Subject: [ADMIN] Procedure for fixing and repair multiple instances
 
Hello,

Please refresh my memory,

I have four instances, configured :

Server AAAA   /db1      instance 1    instance 2
                          /db2      instance 3    instance 4

Postgresql 9.3 running on Debian 7,

My disk array failed and recovered itself, all disk partitions mounted ok, but i have to run fsck in /db1 and /db2,
 I could manually start instance 1 whitout message errors, but instance 2 announces:

       /db1/postgresq9.3/cal/ is not a valid data directory
      file /db1/postgresql9.3/cal/PG_VERSION does not exist

So Instance 1 started ok but Instance 2 does not start, how I fix this error ? I try to re-create this file manually  but later mention to PG_CONTROL does not exist started to appear, this give me the idea that some kind of corruption damaged this directory.

Am I comdenmed to re-install again ???
What could be the best method without touching the instance that is working ok ?

Instance 3 and Instance 4 started ok but several messages like this:
   
        Statistics file pg_stat_tmp/db_12034.stat is corrupted

in their respective log files,

During the fsck processs for /db2 I could see several messages like, specially during phase 2 check directory structure:

ENTRY  'dbo.stat'  which is in /postgresql/9.3/qa/pg_stat_tmp has a incorrect file type (it was 1 but should be 2) Fix(y) ?

if this is related, how I could clean this messages or fix it ?

Thanks a lot

Any idea or recommendation is welcome






Journyx, Inc.
7600 Burnet Road #300
Austin, TX 78757
www.journyx.com

p 512.834.8888 
f 512-834-8858 

Do you receive our promotional emails? You can subscribe or unsubscribe to those emails at http://go.journyx.com/emailPreference/e/4932/714/