hadoop-common-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Andrzej Bialecki ...@getopt.org>
Subject Re: fsck with the move option
Date Tue, 02 May 2006 06:23:25 GMT
Eric Baldeschwieler wrote:
> Why not go after the FS APIs so you can simply read the remaining good 
> blocks?  Might it work already if you had a client that skipped a 
> block if it found a missing one (via error/exception)?  Or create 
> empty blocks...

Well, this version of fsck is an interim solution anyway. Currently 
corrupted files can't be processed, so we need a way to weed them out, 
and possible recover as much data as possible from the leftovers.

My point is that we have no other way to do it now. If/when the missing 
block handling improves so that it doesn't kill jobs, we can change this 
part of fsck to simply move files; for now I think a middle-ground 
solution (additional -recover option) would be better.

Best regards,
Andrzej Bialecki     <><
 ___. ___ ___ ___ _ _   __________________________________
[__ || __|__/|__||\/|  Information Retrieval, Semantic Web
___|||__||  \|  ||  |  Embedded Unix, System Integration
http://www.sigram.com  Contact: info at sigram dot com

View raw message