From: | Simon Riggs <simon(at)2ndQuadrant(dot)com> |
---|---|
To: | Kevin Grittner <Kevin(dot)Grittner(at)wicourts(dot)gov> |
Cc: | Andres Freund <andres(at)2ndquadrant(dot)com>, Robert Haas <robertmhaas(at)gmail(dot)com>, Florian Pflug <fgp(at)phlo(dot)org>, pgsql-hackers(at)postgresql(dot)org |
Subject: | Re: Catalog/Metadata consistency during changeset extraction from wal |
Date: | 2012-06-26 12:35:47 |
Message-ID: | CA+U5nML6-7feA10O3C+A-Nre2PZuf7m++RJUN5mf9PHPDi6FRQ@mail.gmail.com |
Views: | Raw Message | Whole Thread | Download mbox | Resend email |
Thread: | |
Lists: | pgsql-hackers |
On 25 June 2012 17:42, Kevin Grittner <Kevin(dot)Grittner(at)wicourts(dot)gov> wrote:
> This is clearly going to depend on the topology. You would
> definitely want to try to replicate the DDL for the case on which
> Simon is focused (which seems to me to be essentially physical
> replication of catalogs with logical replication of data changes
> from any machine to all others).
Just to remove any doubt: I'm not trying to support a single use case.
The overall proposals include a variety of design patterns. Each of
those covers many reasons for doing it, but end up with same
architecture.
1) Single master replication, with options not possible with physical
2) Multimaster
3) Many to One: data aggregation
4) Online upgrade
I don't think it will be possible to support all of those in one
release. Each has different challenges.
3 and 4 will not be worked on until 9.4, unless someone else is
willing to work on them. That isn't meant to be harsh, just an
explanation of practical reality that I hope people can accept without
needing to argue it.
--
Simon Riggs http://www.2ndQuadrant.com/
PostgreSQL Development, 24x7 Support, Training & Services
From | Date | Subject | |
---|---|---|---|
Next Message | Robert Haas | 2012-06-26 12:40:30 | Re: patch: avoid heavyweight locking on hash metapage |
Previous Message | Satoshi Nagayasu | 2012-06-26 12:11:33 | Re: pg_stat_lwlocks view - lwlocks statistics |