All of lore.kernel.org
 help / color / mirror / Atom feed
From: Ian Zink <zforce@networkusa.net>
To: linux-kernel@vger.kernel.org
Subject: ReiserFS freeze and corruption
Date: 25 Sep 2001 21:46:25 -0500	[thread overview]
Message-ID: <1001472385.784.29.camel@zforce> (raw)

[I am not subscribed to the list please CC me on any response]

About 15minutes ago now, I was merrily gzipping a 370 meg file. I was
checking the progress doing "du file." At about 260 megs I did a "du
file" and the du failed to respond. I tried to ctrl-c du and then gunzip
to no avail. I tried to open a new console to forcibly kill the process.
It wouldn't open. It seemed anything trying to touch the file system
would fail. I then switched to to the text console tried to login from
there, and it froze. Finally, I sucumbed to the fact my system was in
fact going down. I tried to sync the file system alt-sysrq-s. However, I
did not receive the tale-tell "sync" noise. I tried it again... no go. I
tried to flip back into my X session, where in everything just flat out
froze. Upon rebooting (alt-sysrq-b), I received the message reiserfs was
going to roll back its log:

Replay starting
Replay Failure
03:02 rw-1 want=1075572264 limit=6401902

Reiserfs: run fsck.

I managed to find a CD-rescue disk with reiserfs fsck. Upon running it,
it rolled back the journal ( which apparently the kernel couldn't do)
found errors and told me to rebuild the tree. Because I now fimiliar
with the rebuild tree behavior of fsck(that is, you lose random files),
I decided to not go that route... I ran reiserfsck again with the -x
parameter, and it said no errors found. So some how.. the errors
disappeared. Okay. Now everything _seems_ to be intact. However, as
another person pointed out, aparently with reiserfs when things fail you
can random file cross-tamination when reiserfs replays its logs. 

Now for a few details:
Kernel: 2.4.10
gcc version 2.95.4
IBM Deskstar IBM-DTTA-371440
hdparm -X66 -m16 -u1 -c1 -k1 -d1 /dev/hda
Chipset:  Intel Corporation 82371AB PIIX4 IDE
These messages right before the crash:

Sep 25 20:53:13 zforce kernel: attempt to access beyond end of device
Sep 25 20:53:13 zforce kernel: 03:02: rw=1, want=1077688052,
limit=6401902
Sep 25 20:53:22 zforce kernel: attempt to access beyond end of device
Sep 25 20:53:22 zforce kernel: 03:02: rw=1, want=1078220580,
limit=6401902
Sep 25 20:53:42 zforce kernel: attempt to access beyond end of device
Sep 25 20:53:42 zforce kernel: 03:02: rw=1, want=1078289876,
limit=6401902
Sep 25 20:55:04 zforce kernel: attempt to access beyond end of device
Sep 25 20:55:04 zforce kernel: 03:02: rw=1, want=1078444424,
limit=6401902
Sep 25 20:55:26 zforce kernel: attempt to access beyond end of device
Sep 25 20:55:26 zforce kernel: 03:02: rw=1, want=1078470640,
limit=6401902
Sep 25 20:56:20 zforce kernel: attempt to access beyond end of device
Sep 25 20:56:20 zforce kernel: 03:02: rw=1, want=1075572264,
limit=6401902

If you need any more information feel free to contact me.

Thanks,

Ian Zink


                 reply	other threads:[~2001-09-26  2:45 UTC|newest]

Thread overview: [no followups] expand[flat|nested]  mbox.gz  Atom feed

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=1001472385.784.29.camel@zforce \
    --to=zforce@networkusa.net \
    --cc=linux-kernel@vger.kernel.org \
    /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 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.