Xen devel - Jul 2004 - burn in program

Has anyone tried out any burn-in program under xen? I think a neat test would be
something that:
. runs the cpu load up
. pages lots
. ups and downs the network interfaces
. sends and receives lots of packets

running that under, say, 5 domains at once would give a basic test of
functionality

james

> Has anyone tried out any burn-in program under xen? I think a neat test
would be something that:
> . runs the cpu load up
> . pages lots
> . ups and downs the network interfaces
> . sends and receives lots of packets
> 
> running that under, say, 5 domains at once would give a basic test of
functionality
Having some regression testing would be excellent.  We used to do
a nightly build-and-run that stressed dom0 pretty hard by running
and recording the benchmark scores of the following tests, but we
had to stop doing it as we don''t have any spare machines:

 SPEC CPU2000
 PostgreSQL loaded by OSDB 
 ttcp rx/tx using tcp and udp with different MTU sizes
 lmbench
 linux build with -j

If someone wanted to start running some tests it would be very
helpful. It was particularly good at letting us know about
performance bugs, which are always the hardest to find.

Incidentally, we know we''ve currently got a networking rx
performance bug in xeno-unstable at the moment that reduces
rx throughput down to about 500Mb/s. Mark is looking in to it. 

That''s the only outstanding xen/xenlinux bug we know of. Anyone
know different?

Ian





-------------------------------------------------------
This SF.Net email sponsored by Black Hat Briefings & Training.
Attend Black Hat Briefings & Training, Las Vegas July 24-29 - 
digital self defense, top technical experts, no vendor pitches, 
unmatched networking opportunities. Visit www.blackhat.com
_______________________________________________
Xen-devel mailing list
Xen-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/xen-devel

On Jul 13, 2004, at 10:37 PM, James Harper wrote:
> Has anyone tried out any burn-in program under xen? I think a neat 
> test would be something that:
> . runs the cpu load up
> . pages lots
> . ups and downs the network interfaces
> . sends and receives lots of packets
>  
> running that under, say, 5 domains at once would give a basic test of 
> functionality
That''s exactly what I plan to do with CTCS when I finally figure out 
how to get -unstable up and running.   :)

Other than the network stuff, CTCS severely stresses a machine/OS, and 
writing a script to mess with the iface and do some stuff (with iptraf 
perhaps) and hook to the CTCS framework isn''t that difficult. 
It''s not
really a regression test, more a burn-in test, but if a machine passes 
CTCS for 24 hours, in my experience, it''s working just fine, and
it''s a
lot easier to set up and run than a real regression test.

-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-
"I think that''s what they mean by   |
"nickels a day can feed a child."   |       http://www.eff.org/
I thought, "How can food be so      | http://www.anti-dmca.org/
cheap over there?"  It''s not, they  |--------------------------
just eat the nickels." -- Peter Nguyen


-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-
"We all enter this world in the    | Support Electronic Freedom
same way: naked; screaming; soaked |        http://www.eff.org/
in blood. But if you live your     |  http://www.anti-dmca.org/
life right, that kind of thing     |---------------------------
doesn''t have to stop there." -- Dana Gould



-------------------------------------------------------
This SF.Net email sponsored by Black Hat Briefings & Training.
Attend Black Hat Briefings & Training, Las Vegas July 24-29 - 
digital self defense, top technical experts, no vendor pitches, 
unmatched networking opportunities. Visit www.blackhat.com
_______________________________________________
Xen-devel mailing list
Xen-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/xen-devel