The system takes over

Norman Gaywood norm at turing.une.edu.au
Wed Nov 27 19:29:00 CST 2002


On Wed, Nov 27, 2002 at 05:49:03PM -0600, Steve_Boley at Dell.com wrote:
> It's showing that your mirrors are reconstructing, is that happening every
> bootup?  If so then that could be part of the problems and you have a bad
> drive in the system or your raid config is messing up.
> Steve

No it's not doing it everytime. The stuff I showed you was after an
unclean shutdown. The reconstuction is done now. The system was running
quite quickly during reconstuction.

I just tried a copy like this:

(rsh otherhost tar cf - .) | tar xf -

and managed to transfer 13G in about 25 minutes. Great! Just what you
would expect.  The system showed hardly any load at all. Problem is,
I introduced another variable by doing the copy to a disk partition on
a seperate controller to where the SW raid sets are located.

I'm just doing some more copy tests and may have some more clues in a
few hours.

On Thu, Nov 28, 2002 at 09:48:50AM +1000, jason andrade wrote:
> What is the behaviour of the system like without running the rsync process
> over a period of time ?  what does /proc/interrupts show you - are they
> balanced across cpus ?

All good things to look at. The copy tests take several hours, so I'll get
to them soon.

> Are you running the latest errata kernel ?

I'm running:

2.4.18-19.7.tg3.120bigmem

I was running:

2.4.18-17 bigmem from the RH8 updates. That has the same problem while
using the bcm5700 driver. If I use the tg3 driver, the system will
do a "rock-solid-hang" (not pingable) very quickly.

-- 
Norman Gaywood -- School of Mathematical and Computer Sciences
University of New England, Armidale, NSW 2351, Australia
norm at turing.une.edu.au     http://turing.une.edu.au/~norm
Phone: +61 2 6773 2412     Fax: +61 2 6773 3312




More information about the Linux-PowerEdge mailing list