/proc/mdstat bug: 2.6.14.2

/proc/mdstat bug: 2.6.14.2

[Andrew Burgess]

The time and speed display for resync is wrong, the recovery numbers are fine.
The resync is actually running at a few MB/sec.

md1 : active raid6 sdn1[8](S) sde1[9] sdq1[0] sdu1[6] sdo1[5] sdaa3[4] sdab1[2] sds1[1]
      1757815296 blocks level 6, 128k chunk, algorithm 2 [8/6] [UUU_UUU_]
      [>....................]  recovery =  3.6% (10616704/292969216) finish=840.3min speed=5597K/sec
      
md0 : active raid6 sdac2[0] sdz1[4] sdy1[2] sdx1[1] sdw1[3] sdv1[5] sdr2[6]
      1875299328 blocks level 6, 128k chunk, algorithm 2 [8/7] [UUUUUUU_]
      [======>..............]  resync = 33.1% (103563392/312549888) finish=1.5min speed=2288625K/sec

This is a amd64 x2 but running in single processor mode because of all the
timer problems with dual cpus

mdadm - v2.1 - 12 September 2005

Re: /proc/mdstat bug: 2.6.14.2

[Neil Brown]

On Tuesday November 29, aab@cichlid.com wrote:
> The time and speed display for resync is wrong, the recovery numbers are fine.
> The resync is actually running at a few MB/sec.
> 
> md1 : active raid6 sdn1[8](S) sde1[9] sdq1[0] sdu1[6] sdo1[5] sdaa3[4] sdab1[2] sds1[1]
>       1757815296 blocks level 6, 128k chunk, algorithm 2 [8/6] [UUU_UUU_]
>       [>....................]  recovery =  3.6% (10616704/292969216) finish=840.3min speed=5597K/sec
>       
> md0 : active raid6 sdac2[0] sdz1[4] sdy1[2] sdx1[1] sdw1[3] sdv1[5] sdr2[6]
>       1875299328 blocks level 6, 128k chunk, algorithm 2 [8/7] [UUUUUUU_]
>       [======>..............]  resync = 33.1% (103563392/312549888) finish=1.5min speed=2288625K/sec
> 
> This is a amd64 x2 but running in single processor mode because of all the
> timer problems with dual cpus

Weird.....
Does it always report wrong numbers, or was this a once-off?
i.e. if you cat the file 5 times in a row, what sorts of results do
you get?

Are there any interesting kernel messages during the resync?

I'm stumped, so any extra info you might be able to provide will
probably be useful.

Thanks,
NeilBrown

RE: /proc/mdstat bug: 2.6.14.2

[Callahan, Tom]

 
Little off this topic, but how did you get the AMD64 x2 to run in single
processor mode? I was trying to figure this out months ago.

Thanks,
Tom Callahan

-----Original Message-----
From: linux-raid-owner@vger.kernel.org
To: Andrew Burgess
Cc: linux-raid@vger.kernel.org
Sent: 11/29/2005 8:34 PM
Subject: Re: /proc/mdstat bug: 2.6.14.2

On Tuesday November 29, aab@cichlid.com wrote:
> The time and speed display for resync is wrong, the recovery numbers
are fine.
> The resync is actually running at a few MB/sec.
> 
> md1 : active raid6 sdn1[8](S) sde1[9] sdq1[0] sdu1[6] sdo1[5] sdaa3[4]
sdab1[2] sds1[1]
>       1757815296 blocks level 6, 128k chunk, algorithm 2 [8/6]
[UUU_UUU_]
>       [>....................]  recovery =  3.6% (10616704/292969216)
finish=840.3min speed=5597K/sec
>       
> md0 : active raid6 sdac2[0] sdz1[4] sdy1[2] sdx1[1] sdw1[3] sdv1[5]
sdr2[6]
>       1875299328 blocks level 6, 128k chunk, algorithm 2 [8/7]
[UUUUUUU_]
>       [======>..............]  resync = 33.1% (103563392/312549888)
finish=1.5min speed=2288625K/sec
> 
> This is a amd64 x2 but running in single processor mode because of all
the
> timer problems with dual cpus

Weird.....
Does it always report wrong numbers, or was this a once-off?
i.e. if you cat the file 5 times in a row, what sorts of results do
you get?

Are there any interesting kernel messages during the resync?

I'm stumped, so any extra info you might be able to provide will
probably be useful.

Thanks,
NeilBrown
-
To unsubscribe from this list: send the line "unsubscribe linux-raid" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: /proc/mdstat bug: 2.6.14.2

[Andrew Burgess]

>On Tuesday November 29, aab@cichlid.com wrote:
>> The time and speed display for resync is wrong, the recovery numbers are fine.
>> The resync is actually running at a few MB/sec.
>> 
>> md1 : active raid6 sdn1[8](S) sde1[9] sdq1[0] sdu1[6] sdo1[5] sdaa3[4] sdab1[2] sds1[1]
>>       1757815296 blocks level 6, 128k chunk, algorithm 2 [8/6] [UUU_UUU_]
>>       [>....................]  recovery =  3.6% (10616704/292969216) finish=840.3min speed=5597K/sec
>>       
>> md0 : active raid6 sdac2[0] sdz1[4] sdy1[2] sdx1[1] sdw1[3] sdv1[5] sdr2[6]
>>       1875299328 blocks level 6, 128k chunk, algorithm 2 [8/7] [UUUUUUU_]
>>       [======>..............]  resync = 33.1% (103563392/312549888) finish=1.5min speed=2288625K/sec
>> 
>> This is a amd64 x2 but running in single processor mode because of all the
>> timer problems with dual cpus

>Weird.....
>Does it always report wrong numbers, or was this a once-off?

Always, they change slightly each time.

>Are there any interesting kernel messages during the resync?

None that I recall.

>I'm stumped, so any extra info you might be able to provide will
>probably be useful.

Its probably my system, its a new motherboard and I've been having alot of
trouble with it. Its the first time I ever saw a resync and it started by
itself early in the boot so maybe the clock jumped during the boot -- I see
that sometimes.

Thank you for the reply and sorry for the most likely false alarm. It was a
frustrating day...

Re: /proc/mdstat bug: 2.6.14.2

[Luca Berra]

On Tue, Nov 29, 2005 at 11:35:18AM -0800, Andrew Burgess wrote:
>The time and speed display for resync is wrong, the recovery numbers are fine.
>The resync is actually running at a few MB/sec.
>
>md1 : active raid6 sdn1[8](S) sde1[9] sdq1[0] sdu1[6] sdo1[5] sdaa3[4] sdab1[2] sds1[1]
>      1757815296 blocks level 6, 128k chunk, algorithm 2 [8/6] [UUU_UUU_]
>      [>....................]  recovery =  3.6% (10616704/292969216) finish=840.3min speed=5597K/sec
>      
>md0 : active raid6 sdac2[0] sdz1[4] sdy1[2] sdx1[1] sdw1[3] sdv1[5] sdr2[6]
>      1875299328 blocks level 6, 128k chunk, algorithm 2 [8/7] [UUUUUUU_]
>      [======>..............]  resync = 33.1% (103563392/312549888) finish=1.5min speed=2288625K/sec
>
>This is a amd64 x2 but running in single processor mode because of all the
>timer problems with dual cpus

do you have powernow modules loaded on this kernel? they might be
tricking your internal clock.

L.

-- 
Luca Berra -- bluca@comedia.it
        Communication Media & Services S.r.l.
 /"\
 \ /     ASCII RIBBON CAMPAIGN
  X        AGAINST HTML MAIL
 / \

Re: /proc/mdstat bug: 2.6.14.2

[Andrew Burgess]

> Little off this topic, but how did you get the AMD64 x2 to run in single
> processor mode? I was trying to figure this out months ago.

Just boot with maxcpus=1 or make a UP kernel.

I have it working now SMP by dumping 2.6.14 and using 2.6.13.4. Nothing
but trouble with .14 including hosing half of a 3 TB raid6 array. It ran .13
for a day and then the power failed this morning and I accidently let it boot
.14 and it I started getting errors on my 3ware controllers in about an hour.
Back to 13...

Beware if you have 3ware controllers. I have a 7810, 7812 and 8506 in this
machine. They start timing out and dropping disks.