[clug] finding source of high wait states

[steve jenkin]

Have you tried 'top' to help identify where the resources are going??

There is a pernicious problem I once had to deal with on Solaris - rapid
spawning of processes. [with 300Mhz SPARC, we could only manage 50-100
procs/sec. Pretty lousy for CGI performance]

Others on the list might have some good idea of identifying and tracking
that on Linux. [I vaguely recall something about this in a thread a
while back]

Good luck. Would love to know how you track this down.

s

jm wrote on 21/9/09 4:06 PM:
> I've got a machine which is suffering from high CPU wait states. The
> load on the machine is sitting around 15 and it's spending about 60-90%
> of time in the wait state. This is normally assoicated with I/O of some
> form. I'm guessing disk. Is there someway to track down which program(s)
> are causing this, which files are being accessed, is there any i/o
> errors (checked logs including dmesg and don't see any), etc?
> 
> Normally I use iostat, but I came across dstat recently. Here's example
> output for this machine:
> 
> $dstat -D total,sda,sdb,sdc,sdd
> ----total-cpu-usage----
> -dsk/total----dsk/sda-----dsk/sdb-----dsk/sdc-----dsk/sdd-- -net/total-
> ---paging-- ---system--
> usr sys idl wai hiq siq| read  writ: read  writ: read  writ: read  writ:
> read  writ| recv  send|  in   out | int   csw 15   3  20  61   0  
> 1|3118k 3957k:3112k  654k:6684B 3303k: 105B   79B:  19B   79B|   0     0

<snip>

> Jeff.
> 


-- 
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