Yes, it's a shared-disk failover solution as described in [1] and only
node will run the pg instance. We got it fixed by setting ping-interval
in resin from 60s to 0s. It's more like validate every connection before
giving to app to process.
> That's talking about IPv6. Are you using IPv6? Is sun cluster doing
something magic WRT your network?
Well we see this error on a stand-alone sol 10 box with no cluster. How
do I check if it's setup to use IPv6 or not.
Thanks,
Stalin
-----Original Message-----
From: Markus Schiltknecht [mailto:markus@bluegap.ch]
Sent: Friday, September 07, 2007 1:24 AM
To: Subbiah Stalin-XCGF84
Cc: pgsql-admin@postgresql.org
Subject: Re: [ADMIN] Postgres with Sun Cluster HA/Solaris 10
Hi,
Subbiah Stalin-XCGF84 wrote:
> Any pointers to debug this futher is greatly appreciated.
I'm not quite sure, how sun cluster is working, but to me it sounds like
a shared-disk failover solution (see [1] for more details).
As such, only one node should run a postgres instance at any time. Does
sun cluster take care of that?
> "LOG: could not bind IPv6 socket: Cannot assign requested address
> HINT: Is another postmaster already running on port 5432? If not,
> wait a few seconds and retry."
That's talking about IPv6. Are you using IPv6? Is sun cluster doing
something magic WRT your network?
(BTW, please don't cross post (removed -performance). And don't reply to
emails when you intend to start a new thread, thanks.)
Regards
Markus
[1]: Postgres Documentation High Availability and Load Balancing
http://www.postgresql.org/docs/8.2/static/high-availability.html