Re: RAID6, failed device, unresponsive system?

Started by Mathias Burén on Jan 18, 2012

open the first message on its own page

Message details and related content
On 18 January 2012 06:54, Stefan /*St0fF*/ Hübner
[off-list ref] wrote:
Hi

Am 17.01.2012 13:33, schrieb Mathias Burén:
quoted
On 17 January 2012 12:02, Peter Grandi [off-list ref] wrote:
quoted
[ ... ]
quoted
quoted
Why is the system unresponsive, shouldn't it still be OK
after a drive failure?
There is a bit of a difference between a "drive failure" and
some/several bad sectors on a drive.

It is also to wonder whether the partially defective drive has
been "failed" and "removed" from the MD set and perhaps
"deleted" using '/sys/block/sdb/device/delete'.
quoted
Hm, I'm seeing this in dmesg, could it be related? (ioctl lock)
quoted
[425480.928740] md/raid:md0: read error corrected (8 sectors at
223617240 on sdb1)
Note the "read error corrected" (*corrected*) and that is is "8
sectors" may indicate it is one of the drives with 4096B sectors
that is configured as if it has 512B ones.
Right, that is how WD20EARS react.
quoted
quoted
[ ... ]

Overall it is likely that you have just implicitly discovered
how important short settings for Error Recovery Control are, and
to choose drives that allow you to set them:

 http://www.sabi.co.uk/blog/1103Mar.html#110331
--
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
 $ sudo smartctl -l scterc,20,20 /dev/sdb
smartctl 5.42 2011-10-20 r3458 [x86_64-linux-3.2.0-2-ARCH] (local build)
Copyright (C) 2002-11 by Bruce Allen, http://smartmontools.sourceforge.net

Warning: device does not support SCT Error Recovery Control command

:-/ and yes it's a "4KB" drive, a WD20EARS. It failed after almost
11000 hours. Thanks, now I know the reason for the system hang.
Those are not very suited for RAID.  They were the cheapest WD 2TB
drives in the consumer segment, they don't support TLER/ERC.  And from
my experience the replacement drives won't last very long, either.  At
least you have raid6 there...
I know. I picked them because they were cheap, I got 5 of them new for
about 380 USD.
Is the drive that corrects its sectors still in the array (I'd guess
that)?  If yes, just issue the next rma, error is "drive reacts very
slowly".  I fear you have to wait for the first resync or do a ddrescue
with the disk that is still in the array while the array is taken
offline (that way you don't take the chance of another drive failing
while resyncing).

All the best,
Stefan
quoted
(cc Linux RAID)
The drive is now out of the array. I had to pull the power to the
system, physically pull the disk, then boot into single user mode.
There I had to do a force assemble (because the array wouldn't
assemble in a not-clean state automatically). That worked fine, so I
did an fsck which turned out OK, and now it's "checking" the array. I
should've probably checked the array before the fsck, but oh well.
Check still in progress.

Thanks,
M
--
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
Keyboard shortcuts
hback out one level
jnext message in thread / next series-diff position
kprevious message in thread / previous series-diff position
ldrill in
Escclose help / fold thread tree
?toggle this help