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

Re: How well ext3 will tolerate errors?



On Mon, 22 Apr 2002, Stephen C. Tweedie wrote:

>
> Define corruption.  Do you mean lost data, or bogus data?
>

Both, first one is lost data, ie. case when hardware will told that there
is error. The second one is case when there is memory bit flip-flop or
other error and loop-AES is feeding at least 1024 bytes chunk of garbage
to the ext3-system. This case is the most intresting, because I think
it will cause the worst damage, right?

>
> The journal area is mostly unused unless you crash.  If that happens,
> e2fsck needs to take care of things.
...
> If the error is data corruption, then either parts of the journal will
> be silently ignored (corrupting active metadata on the fs), or you'll
> end up copying garbage to active filesystem metadata.
>

And the metadata will be corrupted (obviously). What then? Is it
recoverable by e2fsck? I guess that the answer would be "it depends" but I
try to figure out if I am shooting myself to toe, foot or head with this
hd - loopAES - ext3 combo in the case of data corruption (bogus data).

>
> Bitmaps, inode tables, directories.
>

What are the changes that these errors are recoverable?


Best Regards, Jani

--
Jani Averbach






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