linux-raid.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: NeilBrown <neilb@suse.de>
To: Brad Campbell <lists2009@fnarfbargle.com>
Cc: linux-raid@vger.kernel.org
Subject: Re: What the heck happened to my array?
Date: Fri, 8 Apr 2011 19:52:38 +1000	[thread overview]
Message-ID: <20110408195238.1e051c40@notabene.brown> (raw)
In-Reply-To: <4D9E6285.8000006@fnarfbargle.com>

On Fri, 08 Apr 2011 09:19:01 +0800 Brad Campbell <lists2009@fnarfbargle.com>
wrote:

> On 05/04/11 14:10, NeilBrown wrote:
> 
> > I would suggest:
> >    copy anything that you need off, just in case - if you can.
> >
> >    Kill the mdadm that is running in the back ground.  This will mean that
> >    if the machine crashes your array will be corrupted, but you are thinking
> >    of rebuilding it any, so that isn't the end of the world.
> >    In /sys/block/md0/md
> >       cat suspend_hi>  suspend_lo
> >       cat component_size>  sync_max
> >
> 
> root@srv:/sys/block/md0/md# cat /proc/mdstat
> Personalities : [linear] [raid0] [raid1] [raid10] [raid6] [raid5] [raid4]
> md0 : active raid6 sdc[0] sdd[6](S) sdl[1](S) sdh[9] sda[8] sde[7] 
> sdg[5] sdb[4] sdf[3] sdm[2]
>        7814078464 blocks super 1.2 level 6, 512k chunk, algorithm 2 
> [10/8] [U_UUUU_UUU]
>        [=================>...]  reshape = 88.2% (861696000/976759808) 
> finish=3713.3min speed=516K/sec
> 
> md2 : active raid5 sdi[0] sdk[3] sdj[1]
>        1465146368 blocks super 1.2 level 5, 64k chunk, algorithm 2 [3/3] 
> [UUU]
> 
> md6 : active raid1 sdp6[0] sdo6[1]
>        821539904 blocks [2/2] [UU]
> 
> md5 : active raid1 sdp5[0] sdo5[1]
>        104864192 blocks [2/2] [UU]
> 
> md4 : active raid1 sdp3[0] sdo3[1]
>        20980800 blocks [2/2] [UU]
> 
> md3 : active raid1 sdp2[0] sdo2[1]
>        8393856 blocks [2/2] [UU]
> 
> md1 : active raid1 sdp1[0] sdo1[1]
>        20980736 blocks [2/2] [UU]
> 
> unused devices: <none>
> root@srv:/sys/block/md0/md# cat component_size > sync_max
> cat: write error: Device or resource busy

Sorry, I should have checked the source code.


   echo max > sync_max

is what you want.
Or just a much bigger number.

> 
> root@srv:/sys/block/md0/md# cat suspend_hi suspend_lo
> 13788774400
> 13788774400

They are the same so that is good - nothing will be suspended.

> 
> root@srv:/sys/block/md0/md# grep . sync_*
> sync_action:reshape
> sync_completed:1723392000 / 1953519616
> sync_force_parallel:0
> sync_max:1723392000
> sync_min:0
> sync_speed:281
> sync_speed_max:200000 (system)
> sync_speed_min:200000 (local)
> 
> So I killed mdadm, then did the cat suspend_hi > suspend_lo.. but as you 
> can see it won't let me change sync_max. The array above reports 
> 516K/sec, but that was just on its way down to 0 on a time based 
> average. It was not moving at all.
> 
> I then tried stopping the array, restarting it with mdadm 3.1.4 which 
> immediately segfaulted and left the array in state resync=DELAYED.
> 
> I issued the above commands again, which succeeded this time but while 
> the array looked good, it was not resyncing :
> root@srv:/sys/block/md0/md# cat /proc/mdstat
> Personalities : [linear] [raid0] [raid1] [raid10] [raid6] [raid5] [raid4]
> md0 : active raid6 sdc[0] sdd[6](S) sdl[1](S) sdh[9] sda[8] sde[7] 
> sdg[5] sdb[4] sdf[3] sdm[2]
>        7814078464 blocks super 1.2 level 6, 512k chunk, algorithm 2 
> [10/8] [U_UUUU_UUU]
>        [=================>...]  reshape = 88.2% (861698048/976759808) 
> finish=30203712.0min speed=0K/sec
> 
> md2 : active raid5 sdi[0] sdk[3] sdj[1]
>        1465146368 blocks super 1.2 level 5, 64k chunk, algorithm 2 [3/3] 
> [UUU]
> 
> md6 : active raid1 sdp6[0] sdo6[1]
>        821539904 blocks [2/2] [UU]
> 
> md5 : active raid1 sdp5[0] sdo5[1]
>        104864192 blocks [2/2] [UU]
> 
> md4 : active raid1 sdp3[0] sdo3[1]
>        20980800 blocks [2/2] [UU]
> 
> md3 : active raid1 sdp2[0] sdo2[1]
>        8393856 blocks [2/2] [UU]
> 
> md1 : active raid1 sdp1[0] sdo1[1]
>        20980736 blocks [2/2] [UU]
> 
> unused devices: <none>
> 
> root@srv:/sys/block/md0/md# grep . sync*
> sync_action:reshape
> sync_completed:1723396096 / 1953519616
> sync_force_parallel:0
> sync_max:976759808
> sync_min:0
> sync_speed:0
> sync_speed_max:200000 (system)
> sync_speed_min:200000 (local)
> 
> I stopped the array and restarted it with mdadm 3.2.1 and it continued 
> along its merry way.
> 
> Not an issue, and I don't much care if it blew something up, but I 
> thought it worthy of a follow up.
> 
> If there is anything you need tested while it's in this state I've got ~ 
> 1000 minutes of resync time left and I'm happy to damage it if requested.

No thank - I think I know what happened.  Main problem is that there is
confusion between 'k' and 'sectors' and there are random other values that
sometimes work (like 'max') and I never remember which is which.  sysfs in md
is a bit of a mess.... one day I hope to completely replace it (with back
compatibility of course...)

Thanks for the feedback.

NeilBrown


> 
> Regards,
> Brad
> --
> 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


  reply	other threads:[~2011-04-08  9:52 UTC|newest]

Thread overview: 12+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2011-04-03 13:32 What the heck happened to my array? (No apparent data loss) Brad Campbell
2011-04-03 15:47 ` Roberto Spadim
2011-04-04  5:59   ` Brad Campbell
2011-04-04 16:49     ` Roberto Spadim
2011-04-05  0:47       ` What the heck happened to my array? Brad Campbell
2011-04-05  6:10         ` NeilBrown
2011-04-05  9:02           ` Brad Campbell
2011-04-05 11:31             ` NeilBrown
2011-04-05 11:47               ` Brad Campbell
2011-04-08  1:19           ` Brad Campbell
2011-04-08  9:52             ` NeilBrown [this message]
2011-04-08 15:27               ` Roberto Spadim

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=20110408195238.1e051c40@notabene.brown \
    --to=neilb@suse.de \
    --cc=linux-raid@vger.kernel.org \
    --cc=lists2009@fnarfbargle.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).