swamped mail relays - suggestions

[Martin Hepworth]

Hi Greg

Do you reject unknown users on the inbound MTA - we 550 reject 70% of
our inbound traffic that way.

(nice to see you on another list other than oxlug)

--
Martin Hepworth
Snr Systems Administrator
Solid State Logic
Tel: +44 (0)1865 842300


Greg Matthews wrote:
> I'm running three mail relays served round-robin by DNS. Over the last
> month these have started to struggle to cope with the load.
>
> I've installed a caching DNS server on one and it may have a small
> effect offsetting the RBL lookups but it's too small to be definite. All
> 3 have a DNS server close by on the local network.
>
> I'm thinking about changing the file system used to hold the mqueue.in
> directory. At the moment, it is ext3. Do people have recommendations for
> this? I'm running with an older redhat kernel so XFS isnt an immediate
> option but ReiserFS is.
>
> At the moment I have around 10000 emails in the incoming queue on each
> machine and typical iostat numbers look like:
>
> Device:    rrqm/s wrqm/s   r/s   w/s  rsec/s  wsec/s    rkB/s    wkB/s avgrq-sz avgqu-sz   await  svctm  %util
> /dev/sda     0.80 621.80 12.00 232.00  104.00 6873.60    52.00  3436.80    28.60   964.72  394.43   3.07  74.80
> /dev/sda1    0.00   0.00  0.00  0.00    0.00    0.00     0.00     0.00     0.00     0.00    0.00   0.00   0.00
> /dev/sda2    0.80 621.80 12.00 232.00  104.00 6873.60    52.00  3436.80    28.60   964.72  394.43   3.07  74.80
> /dev/sdb     0.00 620.40  8.80 233.40   70.40 6873.60    35.20  3436.80    28.67   967.70  398.60   3.13  75.80
> /dev/sdb1    0.00   0.00  0.00  0.00    0.00    0.00     0.00     0.00     0.00     0.00    0.00   0.00   0.00
> /dev/sdb2    0.00 620.40  8.80 233.40   70.40 6873.60    35.20  3436.80    28.67   967.70  398.60   3.13  75.80
>
> I'm not exactly expert on interpreting this but the await numbers look
> high as does the queue size (I've just seen it at 951).
>
> The relays are using md to create a raid1 mirror of two scsi disks,
> could the md be a bottleneck too?
>
> Each box has dual 2.8GHz Xeons with hyperthreading turned on - is HT a
> winner or should I turn it off?
>
> Perhaps I just need more/faster hardware?
>
> Any comments and suggestions most welcome.
>
> GREG
> --
> Greg Matthews           01491 692445
> Head of UNIX/Linux, iTSS Wallingford
>
> ------------------------ MailScanner list ------------------------
> To unsubscribe, email jiscmail at jiscmail.ac.uk with the words:
> 'leave mailscanner' in the body of the email.
> Before posting, read the Wiki (http://wiki.mailscanner.info/) and
> the archives (http://www.jiscmail.ac.uk/lists/mailscanner.html).
>
> Support MailScanner development - buy the book off the website!


--
--
Martin Hepworth
Snr Systems Administrator
Solid State Logic
Tel: +44 (0)1865 842300

**********************************************************************

This email and any files transmitted with it are confidential and
intended solely for the use of the individual or entity to whom they
are addressed. If you have received this email in error please notify
the system manager.

This footnote confirms that this email message has been swept
for the presence of computer viruses and is believed to be clean.

**********************************************************************

------------------------ MailScanner list ------------------------
To unsubscribe, email jiscmail at jiscmail.ac.uk with the words:
'leave mailscanner' in the body of the email.
Before posting, read the Wiki (http://wiki.mailscanner.info/) and
the archives (http://www.jiscmail.ac.uk/lists/mailscanner.html).

Support MailScanner development - buy the book off the website!