[Pgpool-general] Health check

[Luis Alberto Pérez Paz]

Hi everybody,

I'm doing a lot of tests with pgpool because we want to implement it in our
production environment.

My laboratory is:
Computer 1: running pgpool
Computer 2: running postgres
Computer 3: running postgres

I'm simulating a fall in one the psql servers (Computer 1 & 2).

When I stop the postgres service, pgpool works great. My health_check
parameters are:

health_check_timeout = 5
health_check_period = 5

However when I unplugged the network cable (or even if I unplugged the power
cable) the health_check_timeout does not work, pgpool delays between 1 and 2
minutes to detect the error.

2008-03-11 13:18:26 LOG:   pid 23451: starting health checking
2008-03-11 13:19:14 ERROR: pid 23451: connect_inet_domain_socket: connect()
failed: No route to host


Is it normal??? Can I detect these errors in another way?


Thanks in advance for your help.


Best Regards,


Luis Alberto Perez Paz

-- 
paz, amor y comprensión
(1967-1994)
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://pgfoundry.org/pipermail/pgpool-general/attachments/20080311/485cc60f/attachment.html