[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

ext2 strange behaviour


A box which has been up for 140 days runs on a single 20GB '/' ext2
partition (as was provided by our colocater) . For long this box has
been untouched but was once used for 'heavy-duty' disk activities in the
past.  when I logged in recently I realised that I'm unable to write to
the disk (even as root).

 # touch test.file
 touch: creating `test': Read-only file system

 # pwd
 # mount
 /dev/hda3 on / type ext2 (rw,errors=remount-ro,errors=remount-ro)
 proc on /proc type proc (rw)
 devpts on /dev/pts type devpts (rw,gid=5,mode=620)

 # cat /etc/debian_version

It is worth noting the following:

  * During those 'heavy-duty' days, the system load reached upto 7.00
    and such. 
  * The filesystem has about 100 million small files on it.  there is
    one such directory with _lot_ of files:

    # time ls | head

    real    12m8.965s
    user    0m2.020s
    sys     12m1.850s

    Notice that the file list hasn't been done yet because there are
    _so_many_ files on it and all the files are under one single parent
    and I think thats the same reason why the kernel takes _all_ the
    time trying to read through its nodes.
  * This box is not physically available. It is a remote colocated box
    (and hence any solution, that you think of would be appreciated if it
    can be applied to a remote machine)

Can physical damages to disks cause such problems?

if in case a fsck has to be run, will it be possible to create a cramfs
with ssh enabled and allow a remote login to the system and fsck this
partition? Are there any ready-made tools/packages available for the

  Thanks in advance,

| If you can shine in radiance like the face of my lady,               |
| O dear moon, I shall hail thee or else bless our love                |
| (extolling beautys best - 8), Thirukkural                            |

Reply to: