Re: Replication docs update

On Wed, Mar 21, 2012 at 05:06:27PM +0100, Magnus Hagander wrote:
> On Wed, Mar 21, 2012 at 16:44, Bruce Momjian <bruce(at)momjian(dot)us> wrote:
> > On Wed, Mar 21, 2012 at 12:01:34PM +0100, Magnus Hagander wrote:
> >> On Tue, Mar 20, 2012 at 22:26, Bruce Momjian <bruce(at)momjian(dot)us> wrote:
> >> > Our chapter, "Comparison of Different Solutions", needs an update to use
> >> > our new streaming replication terminology, and an update to mention the
> >> > synchronous option.
> >> >
> >> > Patch attached.  I would like to apply it to head and 9.1.
> >>
> >> Is it really a good idea to remove the name "hot standby" where you've
> >> done so? It's a term that's pretty well set by now. Maybe instad "hot
> >> standby using transaction log replication" instead of taking it away
> >> completely?
> >
> > Well, hot/warm standby is really a side-feature of replication, not a
> > replication technology itself.  WAL streaming is a replication
>
> True. I'm just saying it's a term that many are familiar with, and
> thus removing it doesn't entirely help. Removing PITR is a good idea
> however :-)

Well, the term doesn't help describe what it is, so familiar doesn't
help us much.

> > technology.  For example, Shared Disk Failover is technically a warm
> > standby too.
>
> Pretty sure Shared Disk qualifies as cold standby.

Why is it cold? It can start up right away. What does warm mean then?

--
Bruce Momjian <bruce(at)momjian(dot)us> http://momjian.us
EnterpriseDB http://enterprisedb.com

+ It's impossible for everything to be true. +