[imp] Server Farms..

[Magnus Nordseth]

Oliver Kuhl:
> First of all I cannot tell you something about real loadbalancers. But I 
> can tell you something about our setup, which works via dns-round-robin. 

The greatest advantages in choosing a loadbalancer over dns-round-robin is
the ability to do healthchecking of the realservers. If one server fails,
it gets automaticly removed from the server pool.

At our site we use linux virtualserver
(http://www.linuxvirtualserver.org). You should check it out.

> Our imp-servers are quite the same hardware. So I have a testing-machine 
> with a little shell-script to sync all necessary files with the 
> production machines using rsync.

We use rdist for distribution of config files.

> >3. What about logfiles - We would have all users mail etc on an NFS
> >share - Can you do the same for logfiles?(Or do you get locking issues?)
> >- From a statistical aspect, it would be a pain to have to combine each
> >"real" servers logfiles, then run analysis. 

A simple merge-sort will do this. I can mail the script we use if you are
interested. You can also check out mod_log_spread, a apache module for
distributing logs in realtime.

> > Also from a support
> >perspective - How are support personnel supposed to know which "real"
> >server a client would actually be connecting to in order to see if they
> >are entering a wrong username/pass etc?

Make horde syslog to a central loghost.

> >4. This is OT, but I thought I would ask anyway - Once we have a "real"
> >server setup they way we want - What imaging software can be used (Quick
> >recovery in the event of server failure (Eg. If Total rebuild is
> >necessary), and also simple for addition of new "real" server)?. I would
> >love a utility that can create a bootable CD from our SOE, giving us the
> >ability to just place this CD into the new server, power up and in a
> >couple of minutes we have a fully functional server ready to be placed
> >into the server farm.

We use pxe and redhat's kickstart to install new servers. Combined with
rdist (which distributes configfiles) a new server is up and running in
about 5 minutes.
 
> Ah, I forgot something to mention: Users get distributed to one of the 
> imp-servers by dns-round-robin, as mentioned above. But they will be 
> redirected to one of the servers permanent before login. So we don't 
> have a session-problem here. Disadvantage: I a server fails a user has 
> to login again. Advantage: Much cheaper than a hardware-loadbalancer.
> 
> So you have to decide, how available your webmail has to be.

A software loadbalancer is a really good alternative. It doesn't cost much,
and is much more configurable than one in harware.


-- 
Magnus Nordseth
echo '[q]sa[ln0=aln256%Pln256/snlbx]sb3135071790101768542287578439snlbxq' | dc