--On Montag, Oktober 06, 2008 18:01:32 -0400 Christopher Browne 
<cbbrowne at ca.afilias.info> wrote:

> For instance, in v2.0, sequence updates only take place when their
> values actually change, which should greatly reduce the cost of having
> many sequences replicated, particularly if only a few are heavily
> used.
>

As far as i know it will involve only a handful tables (with a less number 
of sequences), so this won't be an issue even in 1.2 i think.

> If you have a system that is relatively lightly loaded, then it may be
> OK if replication is *relatively* pretty expensive, as long as it's
> not actually too expensive for your resources to cope with.

Yeah, we are going to propose a test setup for know, we'll see.

Do you think it's worth to add a comment like your's above to the 
documentation, something along the line

"If you can assure your network resources are sufficient, it's likely to 
create clusters with much more nodes."

People likely get confused, since they interpret the current phrasing as a 
Slony-I limit. I'll can prepare a documentation patch, if you agree.

-- 
  Thanks

                    Bernd