John Moran <johnfrederickmoran at gmail.com> writes:

> 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?

What you *particularly* want to make sure of is that you're generating
SYNCs against the "master" any time it's up and running.

The case that turns out notably badly is where the "master" is accepting
changes while slon processes are down, and when slons finally return, it
generates one giant weekend-long SYNC that draws a barrel of changes in
as one big increment.

If:
 a) You keep the slon running against the master, or
 b) You run tools/generate_syncs.sh against the master via a cron job
then you're protected against that pathology.

So, if parts of the system are out for a few hours, it may take a number
of SYNCs to get up to date; things play more nicely if those SYNCs are
small, as opposed to being giant mudballs of changes that take a long
time to apply.
-- 
let name="cbbrowne" and tld="ca.afilias.info" in name ^ "@" ^ tld;;
Christopher Browne
"Bother,"  said Pooh,  "Eeyore, ready  two photon  torpedoes  and lock
phasers on the Heffalump, Piglet, meet me in transporter room three"