backend closed the channel unexpectedly everytime!

Greetings,

It's the first time I face this situation, but my database can't let me
connect anyway. How can I proceed to recover information stored there? Its a
very important database and I need to recover information at any cost. I'm
still on PostgreSQL 7.1.2 with RedHat 6.2. It's an embeded system so I can't
upgrade to a newer PostgreSQL release at this time.

There wasn't reboots or system hangups when problem ocurred for the first
time.

Here is some information about connection attempts.

$ pg_ctl start
postmaster successfully started
$ DEBUG: database system was shut down at 2003-03-28 15:25:38 BRT
DEBUG: CheckPoint record at (7, 479465352)
DEBUG: Redo record at (7, 479465352); Undo record at (0, 0); Shutdown TRUE
DEBUG: NextTransactionId: 9026840; NextOid: 5484180
DEBUG: database system is in production state

$
$ psql template1
Server process (pid 30916) exited with status 11 at Fri Mar 28 15:25:59 2003
Terminating any active server processes...
Server processes were terminated at Fri Mar 28 15:25:59 2003
Reinitializing shared memory and semaphores
pqReadData() -- backend closed the channel unexpectedly.
This probably means the backend terminated abnormally
before or while processing the request.
connection to server was lost
$ DEBUG: database system was interrupted at 2003-03-28 15:25:48 BRT
DEBUG: CheckPoint record at (7, 479465352)
DEBUG: Redo record at (7, 479465352); Undo record at (0, 0); Shutdown TRUE
DEBUG: NextTransactionId: 9026840; NextOid: 5484180
DEBUG: database system was not properly shut down; automatic recovery in
progress...
DEBUG: ReadRecord: record with zero len at (7, 479465416)
DEBUG: redo is not required
DEBUG: database system is in production state
Server process (pid 31844) exited with status 11 at Fri Mar 28 15:26:04 2003
Terminating any active server processes...
Server processes were terminated at Fri Mar 28 15:26:04 2003
Reinitializing shared memory and semaphores
DEBUG: database system was interrupted at 2003-03-28 15:26:01 BRT
DEBUG: CheckPoint record at (7, 479465416)
DEBUG: Redo record at (7, 479465416); Undo record at (0, 0); Shutdown TRUE
DEBUG: NextTransactionId: 9026840; NextOid: 5484180
DEBUG: database system was not properly shut down; automatic recovery in
progress...
DEBUG: ReadRecord: record with zero len at (7, 479465480)
DEBUG: redo is not required
$ DEBUG: database system is in production state
$

$ pg_ctl stop
Smart Shutdown request at Fri Mar 28 15:39:13 2003
DEBUG: shutting down
waiting for postmaster to shut down.....DEBUG: database system is shut down
done
postmaster successfully shut down

Best regards,

----------------------------------------------------------------------------
----
José Vilson de Mello de Farias
Software Engineer

Dígitro Tecnologia Ltda - www.digitro.com.br
APC - Customer Oriented Applications
E-mail: vilson(dot)farias(at)digitro(dot)com(dot)br
Tel.: +55 48 281 7158
ICQ 11866179