> Running in an LPAR you are somewhat limited in that you can only use Linux
> tools and the HMC displays.  I do not know what monitoring packages exist for
> Linux (besides top), that will give you the information that you (and we)
> need to work this out.  Of course under VM there are a number of choices.

top will show you process data
sard will show you disk statistics providing your vendor ships that feature
(look in proc/partitions and you should see stuff like..
22    64   80043264 hdd 23376 219526 1943138 1722140 14309 266197 2244104 2311340 0 
801060 4033700
)
vmstat will provide paging rate information (especially handy is vmstat 1)

Sendmail defaults to stopping listening based on load average, the value it
picks may be a bit low for the S/390 i/o v cpu.

Look in /etc/sendmail.cf for

# load average at which we just queue messages
#O QueueLA=8

# load average at which we refuse connections
#O RefuseLA=12

remove the comment, set values, and restart sendmail.

The default sendmail tuning is fairly defensive for a small box. You can
materially improve its performance, especially for stuff like mailing list
work by tuning it. There are two favoured techniques for that. #1 is to
get a copy of the book and/or read the sendmail faqs in full, the other
is to replace it with exim, postfix or qmail 8)


Alan

Reply via email to