John Moran <johnfrederickmoran at gmail.com> wrote:
>
> Hello,
> 
> I maintain a relatively small slony-I replication set - tables will
> only ever be hundreds of megabytes, and only then after a long time.
> 
> There seems to be a demand for me to turn off slaves or even the
> master during periods of downtime, when our application isn't in use
> at all (mostly night time, or when some slaves aren't needed). While I
> appreciate that Slony-I's "normal mode of operation is that all nodes
> are available", this hasn't been problematic so far. I suspect that
> the volumes of data replicated by our application are small enough for
> slaves to bring their event lag to zero within a minute or two of
> becoming available again.
> 
> Am I asking for trouble by doing this?

Hey ... are we related?

Anyway, I don't know about asking for trouble.  Obviously, the queued
data for replication is going to cause the master database to grow
faster than would be expected while the slaves are off, and there's
going to be a spike in CPU/network/disk activity when you turn them
back on, but if all of those things are within your tolerance level,
you'll probably be OK.

Sounds to me like you're trying to solve the problem incorrectly, however.
I assume the concern is $$ saving by powering off the servers, and I'll
tell you from 1st hand experience that it's probably not that much money.
We're in the midst of management fighting us to save $$ whereever possible
right now, and I just yesterday did a test that demonstrates that going to
the trouble to turn off individual workstations saves us a mere $35 per
year per workstation.  Of course, servers tend to be more power hungry,
so you may need to double to quadruple that value, but (really) how much
money is that?

Perhaps considering some sort of server consolidation would be wiser, such
as vitalization or simply putting multiple services on each server.

Anyway, hope that helps.

-- 
Bill Moran
http://www.potentialtech.com