On Wed, 2008-03-19 at 16:56 +0200, veejar wrote:
> My filesystem (UFS-FreeBSD) was crashed and I have lost files from
> PGSQL DATA DIR FOLDER after fsck-utility.
> I have lost files from directory "pg_xlog"
> I reset my pg_xlog.
> And now have such error by starting pgsql-server:
> LOG: database system was interrupted at 2008-03-18 22:29:48 EET
> LOG: checkpoint record is at 3E/6C000020
> LOG: redo record is at 3E/6C000020; undo record is at 3E/6C000020;
> shutdown TRUE
> LOG: next transaction ID: 0/1624775486; next OID: 9528514
> LOG: next MultiXactId: 6643; next MultiXactOffset: 13690
> LOG: database system was not properly shut down; automatic recovery in progress
> LOG: record with zero length at 3E/6C000068
> LOG: redo is not required
> FATAL: could not access status of transaction 6643
> DETAIL: Could not read from file "pg_multixact/offsets/0000" at
> offset 24576: Success.
> LOG: startup process (PID 6267) exited with exit code 1
> LOG: aborting startup due to startup process failure
> LOG: logger shutting down
> Is it possible to lose this problem?
It is possible to recover some data in most situations, but it can take
anywhere from a day or two. My company offers database recovery
services, if you're willing to pay.
PostgreSQL UK 2008 Conference: http://www.postgresql.org.uk
In response to
pgsql-admin by date
|Next:||From: Steve Crawford||Date: 2008-03-21 19:50:10|
|Subject: Re: Postgres Performance|
|Previous:||From: Uwe C. Schroeder||Date: 2008-03-21 04:51:35|
|Subject: Re: exceeds max_fsm_pages|