Date: 03 Sep 2003 23:36:26 +0200 From: arno@heho.snv.jussieu.fr (Arno J. Klaassen) To: freebsd-fs@freebsd.org Subject: very slow fsck on bad disk Message-ID: <wpwucpsgf9.fsf@heho.snv.jussieu.fr>
next in thread | raw e-mail | index | archive | help
hello,
is this normal ? :
- I have a 40G IBM Deskstar IDE disk, about 1-2 years old,
connected to some -stable Pentium-Pro box
- a while ago, I got lots of "disk errors", I decided
to reboot the box, but then it did not recognise the disk
any longer (so I pulled ou the (data) disk)
- ...
- Time passed, I found out that the last backup of that
disk was rather old, and I'd like to spend some time to
get of the disk whatever is still readable
- I manage to find an ASUS-AMD MB, running 5.1-RELEASE
whos BIOS accepts the disk as secondary master
- I could not mount the disk, since "I/O Error" or
something like that
- I started :
fsck_ffs -y -b 32 /dev/ad2s1e
I started this .... "last saturday" and it's still running.
When I look at the dmesg or /var/log messages, my eye
got triggered by this :
Sep 3 23:02:27 tabarnac kernel: ad2: hard error cmd=read fsbn 16405947 status=59 error=40
Sep 3 23:02:32 tabarnac kernel: ad2: hard error cmd=read fsbn 16405948 status=59 error=40
Sep 3 23:02:37 tabarnac kernel: ad2: hard error cmd=read fsbn 16405949 status=59 error=40
Sep 3 23:02:42 tabarnac kernel: ad2: hard error cmd=read fsbn 16405950 status=59 error=40
Sep 3 23:02:47 tabarnac kernel: ad2: hard error cmd=read fsbn 16769183 of 16769183-16769310 status=59 error=40
Sep 3 23:02:52 tabarnac kernel: ad2: hard error cmd=read fsbn 16769183 status=59 error=40
Sep 3 23:02:57 tabarnac kernel: ad2: hard error cmd=read fsbn 16769184 status=59 error=40
Sep 3 23:03:02 tabarnac kernel: ad2: hard error cmd=read fsbn 16769185 status=59 error=40
i.e., for a long time, every five seconds the error says "next block"
then suddenly, it says "no more errors in between blocks 16405950 and
16769183" (i.e. 363233 blocks ....)
and once again 5 seconds later it says "next block bad as well"
Am i wrong or does this smell like "each surface error is queued
for syslog, syslog print is triggered every 5 seconds, progress
on fsck ata-disks is hold until next syslog message is printed"
Thank you very much in advance.
Arno
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?wpwucpsgf9.fsf>