| 1 |
Ian P. Christian wrote: |
| 2 |
>>You could do something similar by NFS mounting maildirs across a cluster. |
| 3 |
> |
| 4 |
> You can create multiple frontends with this - but you'll still have one NFS |
| 5 |
> server, or perhaps multiple NFS servers. You would have to create the logic |
| 6 |
> and system that allowed the frontend IMAP/POP3 server to select the correct |
| 7 |
> backend. Also, due to the fact that cyrus keeps advanced indexes, cyrus |
| 8 |
> can't operate over an NFS share. Courier doens't provide an 'out of the box' |
| 9 |
> method for creating a two-tiered scalable IMAP cluster... as far as I know. |
| 10 |
> |
| 11 |
> I've seen this argument many times, on a mailing list I'm on the argument... |
| 12 |
> sorry, 'discussion' often occours, and it's nearly always courier vs cyrus, |
| 13 |
> and in this list community cyrus usually comes up top. |
| 14 |
> |
| 15 |
> One day I think I'll setup a test system, and actaully run some benchmarks to |
| 16 |
> settle the dispute once and for all ;) |
| 17 |
> |
| 18 |
|
| 19 |
This Cyrus cluster stuff looks like a big pain in the ass for anything |
| 20 |
larger than a few servers if that. Reminds me of the bad old days when |
| 21 |
people wrote POP proxies to get around that fact that mbox was mostly |
| 22 |
unworkable over NFS. I suppose it's a matter of scale and going with |
| 23 |
Cyrus cluster might not be a bad idea if you need a decent sized system |
| 24 |
and don't plan to grow. I see it having issues if your mail grows into a |
| 25 |
truly large system, but that's probably not a problem most people really |
| 26 |
have to worry about. |
| 27 |
|
| 28 |
However your users are in a database or ldap because /etc/passwd just |
| 29 |
doesn't work over 3k users. If you're users are virtual than you can |
| 30 |
hash them a bit. Yeah you need some logic for account creation, but |
| 31 |
after that it's dead simple. |
| 32 |
|
| 33 |
/var/vmail/$1st-letter/$2nd-letter/$username |
| 34 |
|
| 35 |
1-9 and a-f are on NFS mount #1 |
| 36 |
g-n are on NFS #2 |
| 37 |
o-z are on NFS #3 |
| 38 |
|
| 39 |
Hell to be honest you don't even need to hash things, but it does help |
| 40 |
when you get to that million user range. |
| 41 |
|
| 42 |
Have your servers mount up the space and you've completely avoided any |
| 43 |
nonsense about which server has the info you're looking for. That and I |
| 44 |
have more confidence in a dualheaded NFS box than J Random server being |
| 45 |
availible. |
| 46 |
|
| 47 |
In effect we trade some complexity on the database side, and we'd have |
| 48 |
to have a db or lookup table anyway, for simplicity on the backend. That |
| 49 |
seems much less brittle than the Cyrus cluster. |
| 50 |
|
| 51 |
kashani |
| 52 |
-- |
| 53 |
gentoo-server@g.o mailing list |
Archives