Hi thanks for the feedback
>
> How long is a "long long time"? What does vmstat show during
> the injection
> process? Is it one qmail-inject with lots of Bcc:s? Or one
> qmail-inject
> per recipient?
It takes approximately 6 hours for the script to complete, each
message invokes a separate qmail-inject process as the mails are
customised with the persons name / details etc. The concurrency only
seems to hit about 30- 40 while the script is still pumping messages
into qmail-inject.
>
> > Does anyone have any tips on how to analyse the performance
> > bottle-necks .. disk / bandwidth etc ( this is a redhat linux 6.1
> > box) or tips on a better way of doing this.
>
> vmstat is a good place to start. What does it show during the
delivery
> process?
I havent tried vmstat, but top shows cpu not getting much above 20%
and the box doesnt hit and swap so I think memory is ok.
>
> Are you running a local caching DNS server? Can I recommend djdns :>
No Im using our colo's nameserver which is only a hop away , but that
is a good point I hadnt thought of.
>
> Are you using multiple spindes? Can I recommend that you do.
Do you mean separate hard disks for the queue? .. The queue resides
on a RAID5 hardware controlled array.
>
> Are you using a high performance logging system or are you
> using syslog? Can
> I recommend multilog.
>
Yes I am using multilog - syslog was chewing up 90% of cpu time.
>
> Regards.
>