Stuck reshape

DISCUSSION Larkin Lowrey 2 participants 3 messages Nov 22, 2013 – Nov 25, 2013 MD RAID

Share & tools

NeilBrown

2013-11-24 23:43 UTC #1 lore raw

Quoted 6 lines · On Fri, 22 Nov 2013 13:17:52 -0600 Larkin Lowrey <[email protected]>…

I have a reshape that got stuck on what appears to be step 1 of the process.
What can be done? Will a shutdown be recoverable, or should I be prepared to restore from backup?
I can read from the array w/o error.

If it is convenient to take a backup, or a copy of the more important files, that might not be a bad idea. It shouldn't be necessary, but better safe than sorry.

Quoted 4 lines

I initiated the re-shape with the following command (devices 4->5, level 5->6, chunk 512->256):
mdadm --grow /dev/md1 --level=6 --raid-devices=6 --chunk=256

So devices: 4 -> 6 ??

Quoted 17 lines

The output was:
mdadm: level of /dev/md1 changed to raid6
mdadm: Need to backup 3072K of critical section..
The status is:
md1 : active raid6 sdo1[6] sdn1[5] sdd1[2] sdb1[4] sdc1[0] sda1[1]
      2929890816 blocks super 1.2 level 6, 512k chunk, algorithm 18
[6/5] [UUUU_U]
      [>....................]  reshape =  0.0% (512/976630272)
finish=56199369.9min speed=0K/sec
      bitmap: 0/1 pages [0KB], 1048576KB chunk
This process is chewing up a lot of CPU:
 2858 root      20   0    7936   3692    280 R  91.5  0.0  35:23.72 mdadm
 2856 root      20   0       0      0      0 R  24.4  0.0  14:05.14

Quoted 8 lines

md1_raid6
2857 root 20 0 0 0 0 R 12.2 0.0 4:28.68
md1_reshape
(that's 91.5% for mdadm, 24.4% for md1_raid6, and 12.2% for md1_reshape)
All drives are on-line and functioning normally. I did forget to remove the internal bitmap and I also forgot to use an external backup file.

You don't need a backup file when increasing the number of data drives, and recent kernels don't need you to remove the bitmap (and one those that did, it would fail cleanly).
So these aren't problems.

Still, something is clearly wrong.

It should be completely safe to reboot ... but given that I don't know what the reshape is hanging here I cannot promise that the reshape won't hang again after a reboot.

I'll try to reproduce this and see if I can understand what is happening. Meanwhile ... maybe try killing mdadm. That certainly won't hurt and may help.

NeilBrown

Quoted 59 lines

mdadm - v3.2.6 - 25th October 2012
kernel: 3.11.8-200.fc19.x86_64

Syslog is:

[  601.373480] md: bind<sdn1>
[  601.773482] md: bind<sdo1>
[  601.824051] RAID conf printout:
[  601.824058]  --- level:5 rd:4 wd:4
[  601.824062]  disk 0, o:1, dev:sdc1
[  601.824065]  disk 1, o:1, dev:sda1
[  601.824068]  disk 2, o:1, dev:sdd1
[  601.824071]  disk 3, o:1, dev:sdb1
[  601.824073] RAID conf printout:
[  601.824075]  --- level:5 rd:4 wd:4
[  601.824078]  disk 0, o:1, dev:sdc1
[  601.824080]  disk 1, o:1, dev:sda1
[  601.824083]  disk 2, o:1, dev:sdd1
[  601.824085]  disk 3, o:1, dev:sdb1
[  647.692320] md/raid:md1: device sdd1 operational as raid disk 2
[  647.698547] md/raid:md1: device sdb1 operational as raid disk 3
[  647.704787] md/raid:md1: device sdc1 operational as raid disk 0
[  647.710941] md/raid:md1: device sda1 operational as raid disk 1
[  647.718258] md/raid:md1: allocated 5394kB
[  647.752152] md/raid:md1: raid level 6 active with 4 out of 5 devices,
algorithm 18
[  647.760088] RAID conf printout:
[  647.760092]  --- level:6 rd:5 wd:4
[  647.760096]  disk 0, o:1, dev:sdc1
[  647.760100]  disk 1, o:1, dev:sda1
[  647.760103]  disk 2, o:1, dev:sdd1
[  647.760105]  disk 3, o:1, dev:sdb1
[  648.622041] RAID conf printout:
[  648.622049]  --- level:6 rd:6 wd:5
[  648.622053]  disk 0, o:1, dev:sdc1
[  648.622057]  disk 1, o:1, dev:sda1
[  648.622059]  disk 2, o:1, dev:sdd1
[  648.622062]  disk 3, o:1, dev:sdb1
[  648.622065]  disk 4, o:1, dev:sdo1
[  648.622072] RAID conf printout:
[  648.622074]  --- level:6 rd:6 wd:5
[  648.622077]  disk 0, o:1, dev:sdc1
[  648.622079]  disk 1, o:1, dev:sda1
[  648.622082]  disk 2, o:1, dev:sdd1
[  648.622084]  disk 3, o:1, dev:sdb1
[  648.622087]  disk 4, o:1, dev:sdo1
[  648.622089]  disk 5, o:1, dev:sdn1
[  648.622475] md: reshape of RAID array md1
[  648.626832] md: minimum _guaranteed_  speed: 1000 KB/sec/disk.
[  648.633041] md: using maximum available idle IO bandwidth (but not
more than 200000 KB/sec) for reshape.
[  648.643053] md: using 128k window, over a total of 976630272k.

--Larkin
--
To unsubscribe from this list: send the line "unsubscribe linux-raid" in the body of a message to [email protected]
More majordomo info at http://vger.kernel.org/majordomo-info.html

Larkin Lowrey

author

2013-11-25 01:44 UTC #2 lore raw

Quoted 5 lines · NeilBrown

Quoted 3 lines
I initiated the re-shape with the following command (devices 4->5, level 5->6, chunk 512->256):
mdadm --grow /dev/md1 --level=6 --raid-devices=6 --chunk=256
So devices: 4 -> 6 ??

Correct, devices 4->6.

Quoted 22 lines

Quoted 22 lines
The output was:
mdadm: level of /dev/md1 changed to raid6
mdadm: Need to backup 3072K of critical section..
The status is:
md1 : active raid6 sdo1[6] sdn1[5] sdd1[2] sdb1[4] sdc1[0] sda1[1]
      2929890816 blocks super 1.2 level 6, 512k chunk, algorithm 18
[6/5] [UUUU_U]
      [>....................]  reshape =  0.0% (512/976630272)
finish=56199369.9min speed=0K/sec
      bitmap: 0/1 pages [0KB], 1048576KB chunk
This process is chewing up a lot of CPU:
 2858 root      20   0    7936   3692    280 R  91.5  0.0  35:23.72 mdadm
 2856 root      20   0       0      0      0 R  24.4  0.0  14:05.14
md1_raid6
2857 root 20 0 0 0 0 R 12.2 0.0 4:28.68
md1_reshape

Quoted 6 lines

Quoted 4 lines
(that's 91.5% for mdadm, 24.4% for md1_raid6, and 12.2% for md1_reshape)
All drives are on-line and functioning normally. I did forget to remove the internal bitmap and I also forgot to use an external backup file.
You don't need a backup file when increasing the number of data

drives, and

Quoted 1 line

recent kernels don't need you to remove the bitmap (and one those that

did,

Quoted 6 lines

it would fail cleanly).
So these aren't problems.
Still, something is clearly wrong.
It should be completely safe to reboot ... but given that I don't know

what

Quoted 9 lines

the reshape is hanging here I cannot promise that the reshape won't hang again after a reboot.
I'll try to reproduce this and see if I can understand what is happening. Meanwhile ... maybe try killing mdadm. That certainly won't hurt and may help.
NeilBrown

Thank you for your reply. I managed to resolve this issue but hadn't gotten around to replying with an update.

I noticed, via strace, that mdadm was reading from sysfs as fast as it could and was trying to open sync_completed O_RDWR which it was unable to do since it was read-only. Just for fun I changed sync_completed to 0644 and all mdadm did with that file was read from it. This was with mdadm v3.2.6. This 'fix' did not change any aspect of the behavior other than allowing mdadm to read the file w/o error.

I finally decided to 'echo max > /sys/block/md1/md/sync_max'. It had been stuck at 1024. That unstuck the reshape and it completed successfully. Being paranoid, I did a full compare of the most recent backup and no (unexpected) differences were detected.

I did a 'echo t > /proc/sysrq-trigger' while things were stuck but it overran the buffer rather badly.

--Larkin