Monit problem

[Ycrux]

Hi!

I'm facing a DB monitoring problem with "monit-4.9" under linux (2.6.15).
I want to monitor a "postgres" database (v8.4) with the following "monitrc"
configuration:

---------------- >8 snip >8 snip >8 snip >8 snip >8 snip ----------------
set daemon  120
set logfile syslog facility log_daemon
check process postgresql with pidfile /var/pgsql/data/postmaster.pid
 start program = "/etc/rc.d/postgresql start"
 stop program = "/etc/rc.d/postgresql stop"
 if failed host localhost port 5432 then start
 if 5 restarts within 5 cycles then timeout
---------------- >8 snip >8 snip >8 snip >8 snip >8 snip ----------------

But "monit" is unable to start the DB.
I tried to debug it as explained in the FAQ by following these steps:

a) Start monit (I assume that "postgres" is running here):
/etc/rc.d/monit start

b) Stop "postgres"  service:
/etc/rc.d/postgresql stop

c) Run: strace -f -p $(cat /var/run/monit.pid) 2>&1 | tee /tmp/trace.out

d) please find attached the "/tmp/trace.out" (compressed as trace.out.gz)

It seems that "monit" search for "/var/pgsql/data/postmaster.pid". But 
this doesn't exist yet because "postgres" database  is down (step b).


Could you please help on that?

cheers
Y.

trace.out.gz
Description: application/gunzip