All of lore.kernel.org
 help / color / mirror / Atom feed
* very long fsck(weeks) for a small(500GB) ext4 partition.
@ 2013-01-19 16:40 Denis 'GNUtoo' Carikli
  2013-01-22 13:23 ` Jan Kara
  0 siblings, 1 reply; 3+ messages in thread
From: Denis 'GNUtoo' Carikli @ 2013-01-19 16:40 UTC (permalink / raw)
  To: linux-ext4

Hi,

I hope it's the right mailing list.

I was shrinking my ext4 partition with resize2fs when the electricity
went off.
Then I launched an fsck against my partition which is
at /dev/mapper/something because it's an opened LUKS partition.
The partition is a bit less than 500GB in size because it's a 500GB hdd.

Now the problem is that it's checking the filesystem since more than
433 hours according to htop.

since I've forgetten to add -y in the fsck command I've something
mechanical that keep pressing y on the keyboard all the time.

here's what fsck says at the time of writing the email:
File ... (inode #129283, mod time Tue Sep 25 21:09:35 2012) 
  has 3 multiply-claimed block(s), shared with 1 file(s):
/home/gnutoo/networking/SDR/uhd.20121031204636/host/build/docs/doxygen/???/structuhd_1_1not__implemented__error.html 
(inode #14946569, mod time Tue Sep 25 21:09:35 2012)
Clone multiply-claimed blocks<y>? yes

1) Can I stop the fsck and continue it later?
2) I've no idea how much inodes there are in the filesystem but I fear
that it it's only at the beginning of it and will last forever.
3) If I stop, and mount it, maybe I'll be able to copy the data?

e2fsprogs are at version 1.41.14-1ubuntu3.

Denis.

^ permalink raw reply	[flat|nested] 3+ messages in thread

* Re: very long fsck(weeks) for a small(500GB) ext4 partition.
  2013-01-19 16:40 very long fsck(weeks) for a small(500GB) ext4 partition Denis 'GNUtoo' Carikli
@ 2013-01-22 13:23 ` Jan Kara
  2013-01-23 14:59   ` Denis 'GNUtoo' Carikli
  0 siblings, 1 reply; 3+ messages in thread
From: Jan Kara @ 2013-01-22 13:23 UTC (permalink / raw)
  To: Denis 'GNUtoo' Carikli; +Cc: linux-ext4

  Hello,

On Sat 19-01-13 17:40:44, Denis 'GNUtoo' Carikli wrote:
> I was shrinking my ext4 partition with resize2fs when the electricity
> went off.
  Heh, bad luck...

> Then I launched an fsck against my partition which is
> at /dev/mapper/something because it's an opened LUKS partition.
> The partition is a bit less than 500GB in size because it's a 500GB hdd.
  When something really bad happens, as it happened to you, it is a good
idea to spend some more time, find some external hard drive or spare space
elsewhere and backup the corrupted device. When you later come to a
situation like "Oh, I forgot to give -y to fsck", you can just stop it
without thinking twice and you will be glad you spent the extra time with
finding free space.

> Now the problem is that it's checking the filesystem since more than
> 433 hours according to htop.
> 
> since I've forgetten to add -y in the fsck command I've something
> mechanical that keep pressing y on the keyboard all the time.
  ;)

> here's what fsck says at the time of writing the email:
> File ... (inode #129283, mod time Tue Sep 25 21:09:35 2012) 
>   has 3 multiply-claimed block(s), shared with 1 file(s):
> /home/gnutoo/networking/SDR/uhd.20121031204636/host/build/docs/doxygen/???/structuhd_1_1not__implemented__error.html 
> (inode #14946569, mod time Tue Sep 25 21:09:35 2012)
> Clone multiply-claimed blocks<y>? yes
  I suppose the first inode number is increasing right? As you can see you
are now at inode 129283. I estimate your filesystem has about 31250000
inodes (chosen by mke2fs with default settings for your fs size) and although
not all of them are used by far, it will take years before the fsck will
finish. So I think you can terminate it with peace in mind. There's no
point in letting it run.
 
> 1) Can I stop the fsck and continue it later?
  You can run it again (with -y this time). It shouldn't do much harm
AFAIK.

> 2) I've no idea how much inodes there are in the filesystem but I fear
> that it it's only at the beginning of it and will last forever.
  Yup.

> 3) If I stop, and mount it, maybe I'll be able to copy the data?
  Maybe. It's definitely worth a try. Make sure to mount read-only.

> e2fsprogs are at version 1.41.14-1ubuntu3.
  And using newer e2fsprogs may give better results as well.

								Honza
-- 
Jan Kara <jack@suse.cz>
SUSE Labs, CR

^ permalink raw reply	[flat|nested] 3+ messages in thread

* Re: very long fsck(weeks) for a small(500GB) ext4 partition.
  2013-01-22 13:23 ` Jan Kara
@ 2013-01-23 14:59   ` Denis 'GNUtoo' Carikli
  0 siblings, 0 replies; 3+ messages in thread
From: Denis 'GNUtoo' Carikli @ 2013-01-23 14:59 UTC (permalink / raw)
  To: Jan Kara; +Cc: linux-ext4

Thanks a lot for all the advises.

It seems that mounting it work, and I can read some files(I didn't try
all files yet).

Denis.


^ permalink raw reply	[flat|nested] 3+ messages in thread

end of thread, other threads:[~2013-01-23 19:13 UTC | newest]

Thread overview: 3+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2013-01-19 16:40 very long fsck(weeks) for a small(500GB) ext4 partition Denis 'GNUtoo' Carikli
2013-01-22 13:23 ` Jan Kara
2013-01-23 14:59   ` Denis 'GNUtoo' Carikli

This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.