sar -d 1 5 x fixed values

[Vogra]

root@vsurjux5:/bkp2/admin/vsdb/arch #sar -d 1 5
Hi All,

rinning "sar -d 1 5" I saw the avque of 2 disks has fixed values.
What ist it?

HP-UX vsurjux5 B.11.11 U 9000/800 10/04/04

10:30:27 device %busy avque r+w/s blks/s avwait avserv

Average c0t0d0 6.20 22.50 14 302 5.46 5.25
Average c0t1d0 3.60 47.50 6 87 4.91 14.57

Thanx,
Lima.

[Steven E. Protter]

You have some io/wait on these disks. You may be pushing them to hard. What's on them.

Including a more complete data collection script that lets you collect data over time.

SEP

[G. Vrijhoeven]

Hi Lima.

avque Average number of requests outstanding
for the device;

I think you need to get more that 5 results before you can jump to a conclusion.
Check man sa and gather more information.

As long as your disk is not >50% busy your disk is able to handle the requests it gets. If the responce time is to long the only thing you can do is spead the data amongst more disks/ to faster disks.

HTH,

Gideon

[Dietmar Konermann · Accepted solution]

Lima,

as you may have already noticed sar -d shows "nearly idle" disks with an avque of "0.50" while really idle disks are omitted completely.

You see constantly "22.50" and "47.50" for two disks... which quite looks like well-known "0.50" plus some fixed offset. There are some known problems that may mess-up the kernel's internal counters which are used to calculate these statistics.

For 11.11 the installation of PHKL_30515 and PHKL_30510 should fix this. (BTW, after reboot the symptom disappears anyway, but may appear again).

Best regards...
Dietmar.

[Gian Paolo]

Hi,
I have this problem,
http://forums1.itrc.hp.com/service/forums/questionanswer.do?threadId=998045&admit=-682735245+1139387615095+28353475
can you hel me....?

bye

[Gian Paolo]

Hi,
I have this problem,
http://forums1.itrc.hp.com/service/forums/questionanswer.do?threadId=998045&admit=-682735245+1139387615095+28353475
can you help me....?

bye