2.4.24 I/O error breakage

[Daniel Pittman]

On 7 Jul 2004, Alex Bligh wrote:
> --On 06 July 2004 15:38 +0100 "Stephen C. Tweedie" <sct at redhat.com> wrote:
>
>> Right. That has been the standard behaviour of ext2 and ext3 on certain
>> critical corruptions for a long time.  On ext3, it is usually
>> accompanied by log messages about the journal aborting; all future
>> writes get the EROFS fs-is-readonly error.
>>
>> Of course, if this hits the partition with /var on it, your logs stop
>> being recorded too.
>
> Can this happen due to a /single/ corruption? I would have thought I
> would see the controller/drive being reset and a retry or two before
> this happened.

As Stephen correctly points out, "corruption" means "broken filesystem
data" in the case I was talking about, and it can happen anywhere.

You may want to investigate 'netconsole', if you have it in a suitable
kernel, which I found very helpful -- it let me trap the oops without
needing to find a suitable serial cable. So long as your network card is
still working, anyway. :)
      Daniel
-- 
Fortune rarely accompanies anyone to the door.
        -- Balthasar Gracian