Postgresql -- config file change fails to trigger restart


#1

Hi –

I’m using the postgresql cookbook here:


with Chef-solo, on VirtualBox, using Vagrant.
(I’m also using a repmgr cookbook for replication, but I don’t think
that’s the issue.)

In my own cookbook’s attributes/default.rb file I’m including
default[:postgresql][:config][:listen_addresses] = ‘*’

The chef run fails when repmgr tries to connect to the database and fails.

---- Begin output of /usr/bin/repmgr -f /etc/repmgr/repmgr.conf master
register ----
STDOUT:
STDERR: Connection to database failed: could not connect to server:
Connection refused
Is the server running on host “10.0.2.15” and accepting
TCP/IP connections on port 5432?
---- End output of /usr/bin/repmgr -f /etc/repmgr/repmgr.conf master
register ----
Ran /usr/bin/repmgr -f /etc/repmgr/repmgr.conf master register returned 6
The postgresql server is indeed running. However, it is not listening
on 10.0.2.15. (That’s the local interface set up by VirtualBox.)
If I restart (not reload) postgresql then it does what I expect and
listens on 10.0.2.15.

My expectation is that adding the :listen_address = ‘*’ attribute should
trigger a restart.
The Usage documentation states “On server systems, the postgres server
is restarted when a configuration file changes.”

Am I doing something wrong, or how can I work around this issue? Is
there some way to force a restart?

Thanks for your help!

Liam


Liam Kirsher
PGP: http://liam.numenet.com/pgp/