[clug] finding source of high wait states

[steve jenkin]

jm wrote on 22/9/09 1:08 PM:
> Only problems with the excellent suggestions by all is that it's an old
> machine (Fedora Core 4) and load seems to have vapourise all by itself
> :-). I'll keep these suggestions in mind for future reference.
> 
> Jeff.
> ps. I'm guessing to was somehow spam related. It always seems to be.

jeff,

Do the networking people have charts & graphs available for traffic at
the time?

If you've got port 25 (or others) open to the world, then if your mail
program/inetd spawns a child process, you'd get the behaviour you
observed... System is nailed, high % wait (on IO - swap or network)

If they don't have data, time to set something in place before next time
(yes, Virginia, they will strike again).

I've heard of 'ntop', but not used it.

Others recommend good traffic monitoring & post-analysis tools?

Do you need more than RRDtool & ORCA etc?
Or is SNMP a better row to hoe?

s

> 
> Ian Munsie wrote:
>> I've found atop to be quite handy tracking down which processes are
>> using I/O.
>>
>> -I
>>
>>   


-- 
Steve Jenkin, Info Tech, Systems and Design Specialist.
0412 786 915 (+61 412 786 915)
PO Box 48, Kippax ACT 2615, AUSTRALIA

sjenkin at canb.auug.org.au http://members.tip.net.au/~sjenkin