weird negative seconds on checkpoint message.

Hi!
I have the next messege:
LOG: checkpoints are occurring too frequently (-21093 seconds apart)

Postgres 8.3.11 is running on SCO Openserver 5.0.7

and this is my   checkpoint configuration:
# - Checkpoints -

checkpoint_segments = 32        # in logfile segments, min 1, 16MB each
checkpoint_timeout = 15min        # range 30s-1h
checkpoint_completion_target = 0.7    # checkpoint target duration, 0.0
- 1.0
#checkpoint_warning = 30s        # 0 is off


Any idea, why i get negative value????
El contenido de este correo electrónico y sus archivos adjuntos son privados y confidenciales y va dirigido
exclusivamentea su destinatario.  No se autoriza la utilización, retransmisión, diseminación, o cualquier otro uso de
estainformación por un receptor o entidades distintas al destinatario.  Si recibe este correo sin ser el destinatario
sele solicita eliminarlo y hacerlo del conocimiento del emisor. La empresa no se hace responsable de transmisiones o
comunicacionesno autorizadas o emitidas por personas ajenas a sus colaboradores utilizando éste medio electrónico. 

The content of this email and its attached files are private and confidential and intended exclusively for the use of
theindividual or entity to which they are addressed. The retransmission, dissemination, or any other use of this
informationother than by the intended recipient is prohibited.  If you have received this email in error please delete
itand notify the sender.  The company cannot be held liable for unauthorized electronic transmissions or
communications,nor for those emitted by non-company individuals and entities.