> Apologies if this is old ground, but it's new to me. I am jumping into
> the deep end with Slony here.
>
> Question numero uno:
>
> I have a 300GB db of 13 tables (7 small, 6 large) which I am trying to
> replicate for backup purposes. I have installed Slony 1.1.5 on my master
> (PostgreSQL 7.4.4) and slave (8.1.3).
>
> I have set everything up - all looks good to me, I have a subscribe set
> etc., started my slons on both sides, I get appropriate looking debug
> messages etc. But no data appears in my slave.
>
> I can see a postgres COPY process on the master being executed against
> the ip address of the slave, so something is happening. Should I expect
> to see my initial slave data population to be gradual - e.g. a bundle of
> records at a time, or a table at a time, or the whole database in one go?

You'll see the whole thing get copied at one "swell foop," where it will
take on the state of the master as at the point in time at which the first
COPY started, and then it should catch up.

> If the whole db, any guesses as to how long my 300GB might take ;)

Highly dependent on hardware parameters (pointedly: disk and network). 
I'd expect that much data to take the better part of a day to copy...

> Question numero duo:
>
> After approx 16 hours of running this appeared in my slave log last night:
>
> 2006-08-01 01:26:11 BST DEBUG1 cleanupThread:    0.001 seconds for
> cleanupEvent()
> 2006-08-01 01:26:11 BST DEBUG1 cleanupThread:    0.016 seconds for
> delete logs
> 2006-08-01 01:37:26 BST ERROR  remoteWorkerThread_1: copy from stdin on
> local node - PGRES_FATAL_ERROR ERROR:  invalid UTF-8 byte sequence
> detected near byte 0x92
> CONTEXT:  COPY messageblks, line 13, column messageblk: "This is a
> multi-part message in MIME format.
>
> ------=_NextPart_000_0038_01C5D95E.F4FD95A0
> Content-Ty..."
> 2006-08-01 01:37:26 BST WARN   remoteWorkerThread_1: data copy for set 1
> failed
> - sleep 15 seconds
> 2006-08-01 01:37:41 BST DEBUG1 copy_set 1
> 2006-08-01 01:37:41 BST DEBUG1 remoteWorkerThread_1: connected to
> provider DB
> NOTICE:  truncate of "public"."aliases" succeeded
> NOTICE:  truncate of "public"."auto_notifications" succeeded
> NOTICE:  truncate of "public"."auto_replies" succeeded
> NOTICE:  truncate of "public"."config" succeeded
> NOTICE:  truncate of "public"."mailboxes" failed - doing delete
> NOTICE:  truncate of "public"."messageblks" succeeded
> 2006-08-01 01:37:53 BST DEBUG1 cleanupThread:    0.001 seconds for
> cleanupEvent()
> 2006-08-01 01:37:53 BST DEBUG1 cleanupThread:    0.016 seconds for
> delete logs
>
> As you can see "invalid UTF-8 byte sequence" - any ideas why this should
> be, what I should do about it, do I need to reconfigure locale / charset
> support on my slave?

You'll notice, presumably, that it keeps failing and restarting.

The problem is that you have invalid UTF-8 data in your 7.4 system; 8.1
refuses to accept invalid UTF-8 data.

You need to fix that in the 7.4 system before any kind of load into the
8.1 system can work.

Look in the Slony-I source directory for a file called README.Unicode;
that describes the issue in reasonable detail.