pg.ddx.io  pgsql-admin@postgresql.org mailing list archive  
help / color / mirror / Atom feed
Standby server startup question - not broken, just curious
3+ messages / 3 participants
[nested] [flat]

* Standby server startup question - not broken, just curious
@ 2014-12-17 15:40 John Scalia <jayknowsunix@gmail.com>
  2014-12-17 15:49 ` Re: Standby server startup question - not broken, just curious Scott Whitney <scott@journyx.com>
  0 siblings, 1 reply; 3+ messages in thread

From: John Scalia @ 2014-12-17 15:40 UTC (permalink / raw)
  To: pgsql-admin

Hi all,

I've got 3 PostgreSQL servers in a streaming standby configuration: 1 primary and 2 standby. These have been running great for a long time, and they are all running WAL archiving 
as an insurance policy to the stream. My questions relate to when any standby is started up, I always see in the logs that it tries to read a couple of WAL segments that should be 
present. What would happen if these weren't using WAL archiving - that is the primary wasn't sending them segments?  How many segments is a standby usually looking for (I think I 
always see two)? I've also seen that if one of these segments is broken or malformed, that the standby errors on start and often fails to start. Could I just remove the segment 
causing the problem? After all with streaming working aren't WAL archives just superfluous?
--
Jay


-- 
Sent via pgsql-admin mailing list (pgsql-admin@postgresql.org)
To make changes to your subscription:
http://www.postgresql.org/mailpref/pgsql-admin



^ permalink  raw  reply  [nested|flat] 3+ messages in thread

* Re: Standby server startup question - not broken, just curious
  2014-12-17 15:40 Standby server startup question - not broken, just curious John Scalia <jayknowsunix@gmail.com>
@ 2014-12-17 15:49 ` Scott Whitney <scott@journyx.com>
  2014-12-17 16:02   ` Re: Standby server startup question - not broken, just curious Scott Ribe <scott_ribe@elevated-dev.com>
  0 siblings, 1 reply; 3+ messages in thread

From: Scott Whitney @ 2014-12-17 15:49 UTC (permalink / raw)
  To: John Scalia <jayknowsunix@gmail.com>; +Cc: pgsql-admin

My understanding is this: 

Once the standby stops (restarts, HUP, whatever), it needs to connect to the primary and get anything that didn't happen during the down time. 

This is _always_ somewhere in a WAL, hence wal_keep_segments. If this is set sufficiently high, you don't _need_ WAL archiving. It will find the missing pieces from those, replay, and get back into streaming mode. 

In the event that it's missing pieces it needs to replay, you'll need to reinitialize your standby from a base backup and start it back up. 

Someone please correct me if I am wrong. 

----- On Dec 17, 2014, at 9:40 AM, John Scalia <jayknowsunix@gmail.com> wrote: 

> Hi all,

> I've got 3 PostgreSQL servers in a streaming standby configuration: 1 primary
> and 2 standby. These have been running great for a long time, and they are all
> running WAL archiving
> as an insurance policy to the stream. My questions relate to when any standby is
> started up, I always see in the logs that it tries to read a couple of WAL
> segments that should be
> present. What would happen if these weren't using WAL archiving - that is the
> primary wasn't sending them segments? How many segments is a standby usually
> looking for (I think I
> always see two)? I've also seen that if one of these segments is broken or
> malformed, that the standby errors on start and often fails to start. Could I
> just remove the segment
> causing the problem? After all with streaming working aren't WAL archives just
> superfluous?
> --
> Jay

> --
> Sent via pgsql-admin mailing list (pgsql-admin@postgresql.org)
> To make changes to your subscription:
> http://www.postgresql.org/mailpref/pgsql-admin

Journyx, Inc.
7600 Burnet Road #300
Austin, TX 78757
www.journyx.com

p 512.834.8888
f 512.834.8858
Do you receive our promotional emails? You can subscribe or unsubscribe to those emails at http://go.journyx.com/emailPreference/e/4932/714

^ permalink  raw  reply  [nested|flat] 3+ messages in thread

* Re: Standby server startup question - not broken, just curious
  2014-12-17 15:40 Standby server startup question - not broken, just curious John Scalia <jayknowsunix@gmail.com>
  2014-12-17 15:49 ` Re: Standby server startup question - not broken, just curious Scott Whitney <scott@journyx.com>
@ 2014-12-17 16:02   ` Scott Ribe <scott_ribe@elevated-dev.com>
  0 siblings, 0 replies; 3+ messages in thread

From: Scott Ribe @ 2014-12-17 16:02 UTC (permalink / raw)
  To: Scott Whitney <swhitney@journyx.com>; Scott Whitney <scott@journyx.com>; +Cc: John Scalia <jayknowsunix@gmail.com>; pgsql-admin

On Dec 17, 2014, at 8:49 AM, Scott Whitney <scott@journyx.com> wrote:
> 
> This is _always_ somewhere in a WAL, hence wal_keep_segments. If this is set sufficiently high, you don't _need_ WAL archiving. It will find the missing pieces from those, replay, and get back into streaming mode.

In addition, I believe it looks for local WAL copy first, before fetching from primary.

-- 
Scott Ribe
scott_ribe@elevated-dev.com
http://www.elevated-dev.com/
(303) 722-0567 voice






-- 
Sent via pgsql-admin mailing list (pgsql-admin@postgresql.org)
To make changes to your subscription:
http://www.postgresql.org/mailpref/pgsql-admin



^ permalink  raw  reply  [nested|flat] 3+ messages in thread


end of thread, other threads:[~2014-12-17 16:02 UTC | newest]

Thread overview: 3+ messages (download: mbox mbox.gz follow: Atom feed)
-- links below jump to the message on this page --
2014-12-17 15:40 Standby server startup question - not broken, just curious John Scalia <jayknowsunix@gmail.com>
2014-12-17 15:49 ` Scott Whitney <scott@journyx.com>
2014-12-17 16:02   ` Scott Ribe <scott_ribe@elevated-dev.com>

This inbox is served by DDX for PostgreSQL; see mirroring instructions
for how to clone and mirror all data and code used for this inbox