VERY frustrated with FreeBSD/UFS stability - please help or comment...

[Garance A Drosehn]

At 9:19 AM -0700 5/21/07, Gore Jarold wrote:
>(I love FreeBSD.  This is not a troll.  If all you
>have to contribute is "stop spreading FUD" or "write
>some code yourself" please accept my apologies.)

Okay, I can believe that.  No problem.

>So my plea for help is as follows:
>
>a) am I really the only person in the world that moves
>around millions of inodes throughout the day ?  Am I
>the only person in the world that has ever filled up a
>snapshotted FS (or a quota'd FS, for that matter) ?
>Am I the only person in the world that does a mass
>deletion of several hundred thousand inodes several
>times per day ?

I think this is the main issue.  I did have some problems with
snapshots due to running out of disk-space, etc.  But those
were due to errors on my part, and were not "normal" operating
practice.  The few errors I have run into were pretty easy to
avoid, simply by re-arranging my disk partitions and paying a
little more attention to what I was doing.

For instance, I found out that it's bad to automatically do a
snapshot before an installworld, and then do multiple
installworlds in the same day, while forgetting about all those
snapshots building up.  Sooner or later you have to run out of
disk space, and it's really bad to do that in the middle of an
installworld!

But once I realized what the issue was, I switched to doing one
snapshot per week, instead of one per installworld.  I really
had no practical reason for the snapshot-per-installworld, so
it did not bother me to drop that.

I imagine I've had days where I've deleted one hundred-thousand
inodes in a day, but it would be rare.  And it'd be even more
rare that I'd remove several hundred-thousand inodes multiple
times in the same day.  The one time I was doing something like
that, I put the most volatile data on a partition by itself, and
I'd simply 'newfs' the partition instead of going through and
removing each file.  I didn't do that to avoid any bugs, I only
did it because it was quicker, and I also knew it would get me
back to the exact same starting point for my next run.  And since
part of what I was doing was benchmarking various things, it was
important that each run had the exact same starting point.

I think that your usage is just much larger than the "standard"
freebsd user, so you're hitting edge-cases that the rest of us
are blissfully missing.

-- 
Garance Alistair Drosehn     =               drosehn at rpi.edu
Senior Systems Programmer               or   gad at FreeBSD.org
Rensselaer Polytechnic Institute;             Troy, NY;  USA