Re: Postgres replication question :- One master 2 slaves 9.0.10 - Mailing list pgsql-general

From akp geek
Subject Re: Postgres replication question :- One master 2 slaves 9.0.10
Date
Msg-id CACnhOfJm0JmtcbYZ29AFiSajBhtZQp64UuW3pRPWW+jcivB4Lw@mail.gmail.com
Whole thread Raw
In response to Re: Postgres replication question :- One master 2 slaves 9.0.10  (Jaime Casanova <jaime@2ndquadrant.com>)
List pgsql-general
Thanks for all suggestions. based on the constraints I had with network, I could able to set up the warm stand by. I am seeing the following log file

I don't know to how to handle. 

2013-10-03 17:52:00 GMT [27636]: [457-1] user=,db=LOG:  restored log file "00000001000001F600000003" from archive
scp:/archive/00000001000001F600000004: No such file or directory
2013-10-03 17:52:02 GMT [27636]: [458-1] user=,db=LOG:  unexpected pageaddr 1F5/34000000 in log file 502, segment 4, offset 0

Regards



On Thu, Oct 3, 2013 at 12:34 AM, Jaime Casanova <jaime@2ndquadrant.com> wrote:
On Tue, Oct 1, 2013 at 10:48 PM, Sergey Konoplev <gray.ru@gmail.com> wrote:
> On Tue, Oct 1, 2013 at 6:15 PM, Jaime Casanova <jaime@2ndquadrant.com> wrote:
>
>> but it would be a good idea to set hot_standby_feedback to on and
>> max_standby_archive_delay to something larger than 30s
>
> Doesn't replica need a connection to master for hot_standby_feedback?
>

doh! yes, it needs it...

vacuum_defer_cleanup_age it's the one you should set if never do
streaming... but, of course, that is not accurate enough

--
Jaime Casanova         www.2ndQuadrant.com
Professional PostgreSQL: Soporte 24x7 y capacitación
Phone: +593 4 5107566         Cell: +593 987171157

pgsql-general by date:

Previous
From: Jaime Casanova
Date:
Subject: Comparing Oracle vs Postgres Fwd: [ADMIN] Database replication to 50-80 nodes
Next
From: Michael Graham
Date:
Subject: Fighting the autovacuumer (to prevent wraparound)