In response to Geoffrey <lists at serioustechnology.com>:

> Bill Moran wrote:
> > In response to Geoffrey <lists at serioustechnology.com>:
> > 
> >> If I need to restart the postmaster on my slony replication slave, 
> >> should I shut down the slony daemons first?
> > 
> > If you restart the postmaster without restarting the slons, the slons
> > will be disconnected, sleep for a bit, then reconnect and go back to
> > work.
> > 
> > So, no, you don't have to manually restart the slons.
> 
> So I guess this makes me wonder why the slony documents warn you about 
> connections that are not reliable.  What happens in a situation where, 
> say the master node goes off line temporarily?  Don't the daemons keep 
> retrying?

Because you need to carefully investigate your requirements in the
event that an "unreliable" connection is involved.

Keep in mind that Slony was not _designed_ to work over unreliable
connections.  The fact that it does a pretty good job anyway is a
testament to good design.

However, a very busy database will have a hellatius time keeping in
sync over a flakey connection.  Also, the term "unreliable" means
different things to different people: some people consider a connection
merely "unreliable" if it drops less than 20% of packets on a regular
basis, whereas I would call such a connection outright "broken".

So the point is that Slony is _not_ designed to handle flakey connections.
If the amount of flake on your connection is low enough to allow Slony
to keep up with your volume of data change, then count your blessings.

-- 
Bill Moran
Collaborative Fusion Inc.
http://people.collaborativefusion.com/~wmoran/

wmoran at collaborativefusion.com
Phone: 412-422-3463x4023