* Use latest xfs_repair on older file systems
@ 2016-04-08 20:37 Chris M Moser
2016-04-08 20:49 ` Eric Sandeen
0 siblings, 1 reply; 4+ messages in thread
From: Chris M Moser @ 2016-04-08 20:37 UTC (permalink / raw)
To: xfs
[-- Attachment #1.1: Type: text/plain, Size: 1252 bytes --]
Hello,
I have a file system that experienced issues and would not mount. The
system would not boot either as this file system was listed in /etc/fstab.
I booted to a debian image with xfsprogs 3.2.1 and ran xfs_repair
The host OS now boots and the file system mounts but has only a handful of
the original files.
I unmounted and ran xfs_repair 2.9.4 form the host OS a few more times. (I
read somewhere that running xfs_repair multiple times can find more files)
I still see the same handful of files.
A previous thread (http://oss.sgi.com/archives/xfs/2014-06/msg00204.html)
suggests running the latest xfs_repair as it may be able to reattach more
files.
I booted to a fedora 23 live image and updated xfsprogs to 4.5.0. (openSUSE
Tumbleweed has xfsprogs 4.5.0 but I was not able to get their live image
working) When running xfs_repair on the file system it complains: xfs_repair:
V1 inodes unsupported. Please try an older xfsporgs.
Is it reasonable to assume that xfs_repair has done all it can and I should
now use photorec from the testdisk package to recover what I can?
Thank you for your feedback.
-Chris
--
Christopher M Moser
Seagate Technology LLC
Cluster Lab Administrator
NRM Bldg A/A1-3
chris.m.moser@seagate.com
952-402-8269
[-- Attachment #1.2: Type: text/html, Size: 2105 bytes --]
[-- Attachment #2: Type: text/plain, Size: 121 bytes --]
_______________________________________________
xfs mailing list
xfs@oss.sgi.com
http://oss.sgi.com/mailman/listinfo/xfs
^ permalink raw reply [flat|nested] 4+ messages in thread
* Re: Use latest xfs_repair on older file systems
2016-04-08 20:37 Use latest xfs_repair on older file systems Chris M Moser
@ 2016-04-08 20:49 ` Eric Sandeen
2016-04-08 22:27 ` Dave Chinner
0 siblings, 1 reply; 4+ messages in thread
From: Eric Sandeen @ 2016-04-08 20:49 UTC (permalink / raw)
To: xfs
On 4/8/16 3:37 PM, Chris M Moser wrote:
> Hello,
>
> I have a file system that experienced issues and would not mount. The
> system would not boot either as this file system was listed in
> /etc/fstab. I booted to a debian image with xfsprogs 3.2.1 and ran
> xfs_repair
>
> The host OS now boots and the file system mounts but has only a
> handful of the original files. I unmounted and ran xfs_repair 2.9.4
> form the host OS a few more times. (I read somewhere that running
> xfs_repair multiple times can find more files) I still see the same
> handful of files.
>
> A previous thread
> (http://oss.sgi.com/archives/xfs/2014-06/msg00204.html) suggests
> running the latest xfs_repair as it may be able to reattach more
> files.
>
> I booted to a fedora 23 live image and updated xfsprogs to 4.5.0.
> (openSUSE Tumbleweed has xfsprogs 4.5.0 but I was not able to get
> their live image working) When running xfs_repair on the file system
> it complains: xfs_repair: V1 inodes unsupported. Please try an older
> xfsporgs.
>
> Is it reasonable to assume that xfs_repair has done all it can and I
> should now use photorec from the testdisk package to recover what I
> can?
Yes, unfortunately.
In general, using the latest xfs_repair is a good idea. And in general,
a single pass of xfs_repair should be enough.
In some cases, there are features which only newer xfs_repair understands -
but older versions should simply refuse to run in that case.
Obsolete features do occasionally get dropped upstream, as well - witness
your V1 inode message. However, that was likely just some other form of
corruption, seen as V1 inodes, unless this really was a very, very old
filesystem (created prior to 2007), which seems unlikely.
It sounds like you experienced some very severe corruption; if that is the
case, you might start by making sure your storage devices are in decent
shape.
Oh, and look in lost+found, if most files seem to be "gone" - they may
be there.
-Eric
> Thank you for your feedback.
>
> -Chris
>
> --
> Christopher M Moser
>
> Seagate Technology LLC
> Cluster Lab Administrator
> NRM Bldg A/A1-3
> chris.m.moser@seagate.com <mailto:chris.m.moser@seagate.com>
> 952-402-8269
>
>
>
> _______________________________________________
> xfs mailing list
> xfs@oss.sgi.com
> http://oss.sgi.com/mailman/listinfo/xfs
>
_______________________________________________
xfs mailing list
xfs@oss.sgi.com
http://oss.sgi.com/mailman/listinfo/xfs
^ permalink raw reply [flat|nested] 4+ messages in thread
* Re: Use latest xfs_repair on older file systems
2016-04-08 20:49 ` Eric Sandeen
@ 2016-04-08 22:27 ` Dave Chinner
0 siblings, 0 replies; 4+ messages in thread
From: Dave Chinner @ 2016-04-08 22:27 UTC (permalink / raw)
To: Eric Sandeen; +Cc: xfs
On Fri, Apr 08, 2016 at 03:49:17PM -0500, Eric Sandeen wrote:
> Obsolete features do occasionally get dropped upstream, as well - witness
> your V1 inode message. However, that was likely just some other form of
> corruption, seen as V1 inodes, unless this really was a very, very old
> filesystem (created prior to 2007), which seems unlikely.
That error does not come from inode number checking - it is emitted
if the NLINK feature bit is missing from the superblock feature
field.
> On 4/8/16 3:37 PM, Chris M Moser wrote:
> I unmounted and ran xfs_repair 2.9.4 form the host OS a few more
> times.
Is this filesystem was created with 2.9.4, we're talking about a
filesystem that was released:
xfsprogs-2.9.4 (7 Sep 2007)
And, importantly:
xfsprogs-2.9.5 (21 Jan 2008)
- Updated mkfs.xfs defaults.
It was then very next release that sets NLINK by default via:
commit 8d537733f52a642d471f6781f32f306241dd4308
Author: Niv Sardi <xaiki@sgi.com>
Date: Fri Nov 16 05:16:34 2007 +0000
....
-
V2 inodes per default, and move DFL bits to XFS_DFL_SB_VERSION_BITS,
Activate XFS_SB_VERSION_NLINKBIT per default, which will enable V2 INODES.
refactor bits that we want everytime in XFS_DFL_SB_VERSION_BITS.
-
....
So, your host has the very last version of xfsprogs that defaulted
to v1 inodes, and so recent xfsprogs versions will simply refuse to
run on it. It does not imply there is any corruption, however, you
just need to use a version of xfs_repair that supports v1 inodes...
Cheers,
Dave.
--
Dave Chinner
david@fromorbit.com
_______________________________________________
xfs mailing list
xfs@oss.sgi.com
http://oss.sgi.com/mailman/listinfo/xfs
^ permalink raw reply [flat|nested] 4+ messages in thread
* Re: Use latest xfs_repair on older file systems
@ 2016-04-11 17:15 Chris M Moser
0 siblings, 0 replies; 4+ messages in thread
From: Chris M Moser @ 2016-04-11 17:15 UTC (permalink / raw)
To: xfs
[-- Attachment #1.1: Type: text/plain, Size: 1579 bytes --]
Thank you kindly for you responses.
-Chris
On Fri, Apr 8, 2016 at 3:37 PM, Chris M Moser <chris.m.moser@seagate.com>
wrote:
> Hello,
>
> I have a file system that experienced issues and would not mount. The
> system would not boot either as this file system was listed in /etc/fstab.
> I booted to a debian image with xfsprogs 3.2.1 and ran xfs_repair
>
> The host OS now boots and the file system mounts but has only a handful of
> the original files.
> I unmounted and ran xfs_repair 2.9.4 form the host OS a few more times. (I
> read somewhere that running xfs_repair multiple times can find more files)
> I still see the same handful of files.
>
> A previous thread (http://oss.sgi.com/archives/xfs/2014-06/msg00204.html)
> suggests running the latest xfs_repair as it may be able to reattach more
> files.
>
> I booted to a fedora 23 live image and updated xfsprogs to 4.5.0.
> (openSUSE Tumbleweed has xfsprogs 4.5.0 but I was not able to get their
> live image working) When running xfs_repair on the file system it
> complains: xfs_repair: V1 inodes unsupported. Please try an older
> xfsporgs.
>
> Is it reasonable to assume that xfs_repair has done all it can and I
> should now use photorec from the testdisk package to recover what I can?
>
> Thank you for your feedback.
>
> -Chris
>
> --
> Christopher M Moser
>
> Seagate Technology LLC
> Cluster Lab Administrator
> NRM Bldg A/A1-3
> chris.m.moser@seagate.com
> 952-402-8269
>
>
--
Christopher M Moser
Seagate Technology LLC
Cluster Lab Administrator
NRM Bldg A/A1-3
chris.m.moser@seagate.com
952-402-8269
[-- Attachment #1.2: Type: text/html, Size: 3294 bytes --]
[-- Attachment #2: Type: text/plain, Size: 121 bytes --]
_______________________________________________
xfs mailing list
xfs@oss.sgi.com
http://oss.sgi.com/mailman/listinfo/xfs
^ permalink raw reply [flat|nested] 4+ messages in thread
end of thread, other threads:[~2016-04-11 17:15 UTC | newest]
Thread overview: 4+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2016-04-08 20:37 Use latest xfs_repair on older file systems Chris M Moser
2016-04-08 20:49 ` Eric Sandeen
2016-04-08 22:27 ` Dave Chinner
-- strict thread matches above, loose matches on Subject: below --
2016-04-11 17:15 Chris M Moser
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox