[prev in list] [next in list] [prev in thread] [next in thread] 

List:       gfs-devel
Subject:    Re: performance anomaly and bug in rm -rf
From:       Ken Preslan <kpreslan () sistina ! com>
Date:       2000-10-14 14:43:39
[Download RAW message or body]

On Wed, Oct 11, 2000 at 09:54:57PM -0500, Andrew Barry wrote:
> Quotas are nice, and all, assuming they are used. Yeah we can say that
> it's the stupid user's fault for not using the features we give them
> (assuming we add quota support in short order here), but we should really
> have a peice of code that will take those dinodes, and reclaim them so
> they can be used for file data, even if it requires quiesing the fs, or
> taking the cluster off-line. It'd still be better than re-making the fs. 

I've figured out how to add code to GFS that can use a very short hiccup
in GFS operation (similar to in concept and implementation to what happens
during journal recovery) to reclaim those unused metadata blocks.  (We also
think we have a fairly good grasp on how to do fast cluster quotas too.)

The DOS attack that started this thread shouldn't be a problem.  (Once we
get a chance to write the code.  :-)

-- 
Ken Preslan <kpreslan@sistina.com>
Code Monkey
Sistina Software, Inc.


-
To unsubscribe from this list: send the line "unsubscribe gfs-devel" in
the body of a message to majordomo@sistina.com
Read the GFS Howto:  http://www.globalfilesystem.org/howtos/gfs_howto/

[prev in list] [next in list] [prev in thread] [next in thread]