Re: integration of pgcluster into postgresql

Adding -hackers back in...

-----Original Message-----
>From: Chahine Hamila [mailto:chahine(dot)hamila(at)yahoo(dot)com]
>Sent: Fri 8/25/2006 8:36 PM
>To: Jim Nasby
>Subject: Re: [HACKERS] integration of pgcluster into postgresql
>
>> First, you need to review all the past discussion
>> about the very
>> intentional decision not to build any replication
>> into the core
>> database.
>
>I would gladly do so. Can you send me any pointer?

I don't really have any handy, but try searching the hackers archive for 'replication'.

>> Second, pgcluster is (AFAIK) command-based
>> replication, which has some
>> very, very serious drawbacks. If PostgreSQL were to
>> include a
>> replication solution, I'd certainly hope it wouldn't
>> be command-based.
>
>It's better than no replication at all... It's good
>enough for many uses.

As is Slony. And dbmirror. And pgpool. So where do we draw the line? Should we include all four?

Yes, I forgot to include hackers on that mail. Anyway,
relax Jim, I'm not trying to invade anyone's turf
here. There seems to be support for the idea of
providing an interface plug for replication modules,
which is fine with me. If you have any constructive
criticism towards that, I'd be most happy to consider
it and try to find an accomodation.

Best regards

--- Jim Nasby <jnasby(at)pervasive(dot)com> wrote:

> Adding -hackers back in...
>
> -----Original Message-----
> >From: Chahine Hamila
> [mailto:chahine(dot)hamila(at)yahoo(dot)com]
> >Sent: Fri 8/25/2006 8:36 PM
> >To: Jim Nasby
> >Subject: Re: [HACKERS] integration of pgcluster
> into postgresql
> >
> >> First, you need to review all the past discussion
> >> about the very
> >> intentional decision not to build any replication
> >> into the core
> >> database.
> >
> >I would gladly do so. Can you send me any pointer?
>
> I don't really have any handy, but try searching the
> hackers archive for 'replication'.
>
>
> >> Second, pgcluster is (AFAIK) command-based
> >> replication, which has some
> >> very, very serious drawbacks. If PostgreSQL were
> to
> >> include a
> >> replication solution, I'd certainly hope it
> wouldn't
> >> be command-based.
> >
> >It's better than no replication at all... It's good
> >enough for many uses.
>
> As is Slony. And dbmirror. And pgpool. So where do
> we draw the line? Should we include all four?
>
> ---------------------------(end of
> broadcast)---------------------------
> TIP 3: Have you checked our extensive FAQ?
>
> http://www.postgresql.org/docs/faq
>

__________________________________________________
Do You Yahoo!?
Tired of spam? Yahoo! Mail has the best spam protection around
http://mail.yahoo.com

On Tue, Aug 29, 2006 at 07:19:09AM -0700, Chahine Hamila wrote:
> Yes, I forgot to include hackers on that mail. Anyway,
> relax Jim, I'm not trying to invade anyone's turf
> here. There seems to be support for the idea of
> providing an interface plug for replication modules,
> which is fine with me. If you have any constructive
> criticism towards that, I'd be most happy to consider
> it and try to find an accomodation.

Well, the big challenge there is that each replication system uses a
different methodology, so you're unlikely to come up with anything that
would be common between any two systems.

I think the best bet is to look for things that can be added that are
either difficult or impossible to do outside the backend, or that have
use beyond just replication.
--
Jim C. Nasby, Sr. Engineering Consultant jnasby(at)pervasive(dot)com
Pervasive Software http://pervasive.com work: 512-231-6117
vcard: http://jim.nasby.net/pervasive.vcf cell: 512-569-9461