On Wed, 9 Jun 1999, Christian E wrote:
> Hi, all
>
> I'm setting up a Linux box with 2.0.36-3 and with SMP enabled...Does
> anyone have experience with the stabilty of such a setup ??
It depends a bit on your particular hardware, but I've run 2.0.36 SMP
boxes for as much as 172 days (and counting) under load. That's pretty
stable.
However, there is very little reason at this point to install 2.0.36
SMP. 2.2.x is also stable on most platforms and is a superior kernel.
All the major commercial distributions now come with 2.2.x IIRC; 2.2.9
in the case of Red Hat 6.0, and will install the SMP kernel
"automagically" if they detect that your system is smp during the
install. I've had only a bit of trouble installing 2.2.4 on a RH 5.2
system to replace the 2.0.36 kernel, but I'd strongly urge getting a
2.2.x-ready distribution and upgrading because there are some tricky
aspects of 2.2, kerneld, and required support libraries (and some bugs
to overcome).
I haven't been able to accumulate the awesome statistics yet for 2.2.x
kernels that I have for 2.0.36 simply because we're awaiting the "right
moment" for an upgrade here (namely, when nobody is doing anything
important) and my 2.2.x UP systems keep getting rebooted by my kids at
home. However, we installed RH 6.0 out of the box on a beowulf made of
Netfinity duals at Linux Expo and it worked "flawlessly"; the only
hassles we experienced were a not unreasonable need to rebuild things
like perl and pvm after the install -- 5.2-developed RPM's didn't work
quite right for reasons that it wasn't worth figuring out at the time
but may well have been our fault anyway.
2.2.x is better overall, handles IRQ assignments and spinlocks MUCH
better, is the supported kernel at this point (2.0.36 is basically
frozen, and bugs you encounter -- however rare they may be -- may never
get fixed). There are also a lot more and better and cooler devices in
2.2.x, many of which will never appear in 2.0.36. Go with 2.2.x.
rgb
>
> best regards
>
> Christian E
> -
> Linux SMP list: FIRST see FAQ at http://www.irisa.fr/prive/mentre/smp-faq/
> To Unsubscribe: send "unsubscribe linux-smp" to [EMAIL PROTECTED]
>
Robert G. Brown http://www.phy.duke.edu/~rgb/
Duke University Dept. of Physics, Box 90305
Durham, N.C. 27708-0305
Phone: 1-919-660-2567 Fax: 919-660-2525 email:[EMAIL PROTECTED]
-
Linux SMP list: FIRST see FAQ at http://www.irisa.fr/prive/mentre/smp-faq/
To Unsubscribe: send "unsubscribe linux-smp" to [EMAIL PROTECTED]
