[Eug-lug] Gentoo hardened

[Bob Miller]

larry price wrote:

> > Heh, 2^10 small reads and writes, then pull the plug.  How long did the
> > reads and writes take, and did the filesystem get corrupted.
> >
> Or N reads/writes/close  on /mnt/tmp;  umount /mnt/tmp; mount
> /mnt/tmp; check files

This completely misses what Jake is talking about.

One major difference between *BSD and Linux is the filesystem.
Linux's ext3 uses much more aggressive write caching than BSD's FFS.
The BSD guys say that their file system is more robust in system
crashes.

That's why Jake wants to pull the plug - to force a system crash,
and demonstrate (he hopes) that Linux loses more files than BSD.
(I personally am skeptical based on my own experience with ext3,
but it would be good to see what Jake's test shows.)

The umount/mount test won't test robustness because umount forces
a sync before it completes.

> well, I'll ask around and see if we can get OPN to loan us a low-end
> box and just do one OS a week for several weeks, on the premise that 
> actually hitting the limits of a slower older box will tell us more
> about the overhead imposed by the OS.

I don't buy that.  As systems architecture evolves, performance
doesn't scale homogeneously.  For example, new systems have bigger
caches, more RAM, and a greater disparity between memory speed and
disk speed and between CPU speed and memory speed.  New CPUs will also
(in general) take more clocks to dispatch an exception, because
there's more processor state to save and restore.  Each of those will
skew a system-level benchmark -- a benchmark that's 15% disk bound
on an i486 might be 80% disk bound on a Pentium 4.

But, since I'm not able to donate a 3GHz system to the cause... (-:

-- 
Bob Miller                              K<bob>
kbobsoft software consulting
http://kbobsoft.com                     kbob at jogger-egg.com