Servidor no se recupera despues de una caida...

Hola, tengo problemas con mi servidor postgres, ya van dos caídas desde ayer
y no ha podido recuperarse, he tenido que volver a instalar postgres y crear
nuevamente la base de datos con el último respaldo con pérdida de datos en
el tiempo en que estaba funcionando ya que el sistema está en producción.

El servidor está en corriendo bajo Windows, no hay problemas de espacio en
disco, ni velocidad de procesador, lo que si, no he cambiado ningún
parámetro de configuración del postgres, dejé lo que está por default (eso
podria tener que ver en las caidas?) El sistema ha estado funcionando
relativamente desde noviembre y ayer fue la primera caida y hoy en la tarde
la segunda, en las dos no se pudo recuperar.

Revisando el log estos son los mensajes antes de la caida:

2007-01-02 08:19:40 LOG: received fast shutdown request

2007-01-02 08:19:40 LOG: aborting any active transactions

2007-01-02 08:19:40 FATAL: terminating connection due to administrator
command

2007-01-02 08:19:40 FATAL: terminating connection due to administrator
command

2007-01-02 08:19:40 FATAL: terminating connection due to administrator
command

2007-01-02 08:19:40 FATAL: terminating connection due to administrator
command

2007-01-02 08:19:40 FATAL: terminating connection due to administrator
command

2007-01-02 08:19:40 FATAL: terminating connection due to administrator
command

2007-01-02 08:19:40 FATAL: terminating connection due to administrator
command

2007-01-02 08:19:40 FATAL: terminating connection due to administrator
command

2007-01-02 08:19:40 FATAL: terminating connection due to administrator
command

2007-01-02 08:19:40 FATAL: terminating connection due to administrator
command

2007-01-02 08:19:40 FATAL: terminating connection due to administrator
command

2007-01-02 08:19:40 FATAL: terminating connection due to administrator
command

2007-01-02 08:19:40 FATAL: terminating connection due to administrator
command

2007-01-02 08:19:40 FATAL: terminating connection due to administrator
command

2007-01-02 08:19:40 FATAL: terminating connection due to administrator
command

2007-01-02 08:19:57 LOG: checkpoints are occurring too frequently (16
seconds apart)

2007-01-02 08:19:57 HINT: Consider increasing the configuration parameter
"checkpoint_segments".

2007-01-02 08:20:02 FATAL: terminating connection due to administrator
command

2007-01-02 08:20:07 LOG: shutting down

Y al tratar de recuperarse este es el log:

2007-01-02 08:21:32 LOG: database system shutdown was interrupted at
2007-01-02 08:20:07 Hora est. del Pacífico de SA

2007-01-02 08:21:32 LOG: checkpoint record is at 3/7E043F20

2007-01-02 08:21:32 LOG: redo record is at 3/7E037ED8; undo record is at
0/0; shutdown FALSE

2007-01-02 08:21:32 LOG: next transaction ID: 1791243; next OID: 1550932

2007-01-02 08:21:32 LOG: database system was not properly shut down;
automatic recovery in progress

2007-01-02 08:21:32 LOG: redo starts at 3/7E037ED8

2007-01-02 08:21:33 PANIC: could not open relation 1663/18515/1545642: No
such file or directory

This application has requested the Runtime to terminate it in an unusual
way.

Please contact the application's support team for more information.

2007-01-02 08:21:33 LOG: startup process (PID 1492) was terminated by
signal 3

2007-01-02 08:21:33 LOG: aborting startup due to startup process failure

2007-01-02 08:21:33 LOG: logger shutting down

En la segunda caída este fue el log:

2007-01-03 14:46:29 LOG: received fast shutdown request

2007-01-03 14:46:29 LOG: aborting any active transactions

2007-01-03 14:46:29 FATAL: terminating connection due to administrator
command

2007-01-03 14:46:29 FATAL: terminating connection due to administrator
command

2007-01-03 14:46:29 LOG: checkpoints are occurring too frequently (0
seconds apart)

2007-01-03 14:46:29 HINT: Consider increasing the configuration parameter
"checkpoint_segments".

2007-01-03 14:46:29 FATAL: terminating connection due to administrator
command

2007-01-03 14:46:29 FATAL: terminating connection due to administrator
command

2007-01-03 14:46:50 LOG: server process (PID 3480) was terminated by signal
58

2007-01-03 14:46:50 LOG: terminating any other active server processes

2007-01-03 14:46:50 WARNING: terminating connection because of crash of
another server process

2007-01-03 14:46:50 DETAIL: The postmaster has commanded this server
process to roll back the current transaction and exit, because another
server process exited abnormally and possibly corrupted shared memory.

2007-01-03 14:46:50 HINT: In a moment you should be able to reconnect to
the database and repeat your command.

2007-01-03 14:46:51 FATAL: terminating connection due to administrator
command

2007-01-03 14:46:51 DETAIL: The postmaster has commanded this server
process to roll back the current transaction and exit, because another
server process exited abnormally and possibly corrupted shared memory.

2007-01-03 14:46:51 HINT: In a moment you should be able to reconnect to
the database and repeat your command.

2007-01-03 14:46:51 WARNING: terminating connection because of crash of
another server process

2007-01-03 14:46:52 DETAIL: The postmaster has commanded this server
process to roll back the current transaction and exit, because another
server process exited abnormally and possibly corrupted shared memory.

2007-01-03 14:46:52 HINT: In a moment you should be able to reconnect to
the database and repeat your command.

2007-01-03 14:46:52 WARNING: terminating connection because of crash of
another server process

…

2007-01-03 14:47:01 DETAIL: The postmaster has commanded this server
process to roll back the current transaction and exit, because another
server process exited abnormally and possibly corrupted shared memory.

2007-01-03 14:47:01 HINT: In a moment you should be able to reconnect to
the database and repeat your command.

2007-01-03 14:47:08 LOG: all server processes terminated; reinitializing

Si alguien me puede ayudar les agradecería muchísimo, ya que la verdad estoy
muy preocupada porque no puedo asegurar la confiabilidad en la base de
datos.

Att,

Paulina Quezada