Greywolf writes:

> As far as RAID goes, I think some serious stress-testing needs
> to happen (thanks, Herb, for starting on it; perhaps an individual
> with non-critical data could step forward and take over? :-).

Suggestions for stress-testing are welcome.  My test box sees load averages of 
100+ when I'm testing major RAID changes.  (The load mix here is typically 
tarring up directories (to /dev/null), extracting large tar files, copying 
data from one part of the raid set to another, copying the root filesystem on
to the raid set, doing a number of "du"'s, running a number of 'bonnies', 
and whatever else I happen to feel like running at the time (and most of the 
above is time-staggered, so that the memory caches play a very small role).  
The system gets to do this for hours at a time, and depending on what's 
being tested, even gets drives powered off during that time...)

While my testing certainly can't cover every conceivable situation, I'd like
to think that the RAIDframe we currently have in-tree is quite stable, and
has been well-tested.  I certainly intend to maintain at least that level of 
stability with any changes that I make to the RAIDframe code...  If other
people wish to stress test the RAID code, that would be great, but I'd 
like to think they'd only be confirming that it works well... 
[Yes, I believe that RAIDframe is solid, and yes, I run it on my machines,
both at home and at work, and yes I would (and have) recommend it to others.]

Later...

Greg Oster