Re: sard/iostat disk I/O statistics/accounting for 2.5.8-pre3

Zlatko Calusic (zlatko.calusic@iskon.hr)
Fri, 12 Apr 2002 01:06:10 +0200


Mike Fedyk <mfedyk@matchmail.com> writes:

> On Thu, Apr 11, 2002 at 07:20:54PM +0200, Zlatko Calusic wrote:
>> Christoph Hellwig <hch@infradead.org> writes:
>> > I'm unable to reproduce it here - I only have idle partition though..
>>
>> Here is how it looks here (SMP machine, it could matter):
>>
>>
>> major minor #blocks name rio rmerge rsect ruse wio wmerge wsect wuse running use aveq
>>
>> 34 0 40011552 hdg 23 62 189 70 4 3 32 0 -1 946410 -946410
>> 33 0 60051600 hde 8349 18725 214666 108230 3781 15234 152176 91700 -1 927060 -746550
>> 3 0 19938240 hda 848 1023 14565 5470 1303 1542 22768 300 -1 942120 -940710
>>
>>
>> Notice negative numbers for running, aveq. Kernel is 2.4.19-pre5-ac3.
>
> major minor #blocks name rio rmerge rsect ruse wio wmerge wsect wuse running use aveq
>
> 3 0 10037774 ide/host0/bus0/target0/lun0/disc 2381496 3602026 12295120 24796300 2167401 11887863 28592436 168474630 -1 849341880 -666064880
> 3 1 488344 ide/host0/bus0/target0/lun0/part1 53667 86 430048 544050 18970 47327 535456 2558790 0 499660 3105630
> 3 2 9549288 ide/host0/bus0/target0/lun0/part2 2327827 3601934 11865056 24252240 2148431 11840536 28056980 165915840 0 16504720 190278950
>
> 2.4.19-pre4-ac3
>
> Did I misunderstand when you reported that statistics were not reported for
> partitions in a previous message? Also -1 only shows up for the drive and
> not the partitions.
>

I was wrong, per-partition information IS correctly reported for
2.4.x. Only in 2.5.x I had to workaround to get per-partition data.

Also, it _seems_ that only on whole devices ios_in_flight variable
(column "running" in /proc/partitions) drops to -1. Your (and mine)
output shows it clearly, yes. This needs further investigation, I have
been unable to find the problem so far.

And, as aveq value is calculated directly from the ios_in_flight
value, the average queue value is wrong too (it never goes negative,
if everything is all right).

-- 
Zlatko
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/