Second speed test

[Jim Levie]

On Sun, 2002-09-22 at 17:06, Robert A. Thompson wrote:
> > When I slam 15k messages into the input queue I don't get scanning rates
> > nearly as good. I didn't keep the numbers for a single large queue run
> > because I was more interested in determining the performance of what one
> > sees on a real mail server. Perhaps I'll re-run my sample with a monster
> > queue.
>
> just a wild guess shooting from the hip, but I would guess this is b/c
> of the time it takes to run through queue to figure out which ones need
> scanning.  Try running a "time mailq"  or "time sendmail -bp
> -Q/path/to/your/mailq.in" and see how long it takes sendmail to run
> through it(with varying number of msg's in the queue).  This may be more
> a limit of dealing with large number of files in a single directory 15k
> msg's =~ 30k files with small files that is a lot of read head movement
> going from file to file no matter what kind of disk you have.
>
Yeah, Its the time that it takes to walk the queue that's the problem.
I've never bothered to see if the MailScanner code handled large dirs
well because that's not what you typically see in the real world. I
quess my point was that a single large queue run probably doesn't
reflect the real performance of MailScanner that you'd expect to see on
a real MailServer.
--
The instructions said to use Windows 98 or better, so I installed
RedHat.