On Wednesday 05 January 2005 02:22 pm, Robert Olsson wrote:
> Jeremy M. Guthrie writes:
> > After smp_affinity adjustments and turning off IRQ balancing.
> >
> > eth3 Link encap:Ethernet HWaddr 00:02:B3:D5:7E:30
> >
> > RX packets:537271635 errors:2377048 dropped:2377048
> > overruns:1849169
>
> Worse?
>
> Yes. I'll remember Hararld moved rt_cache_stat /proc/net/stats/ and wrote
> a new utility for it. Should be in iproute2 package. Stephen knows the
> details. Otherwise change path in rtstat. You need this to relate and
> verify the load.
Hm...
ls -la /proc/net/stat/
total 0
dr-xr-xr-x 2 root root 0 Jan 5 14:47 .
dr-xr-xr-x 5 root root 0 Jan 5 11:50 ..
-r--r--r-- 1 root root 0 Jan 5 14:47 arp_cache
-r--r--r-- 1 root root 0 Jan 5 14:47 clip_arp_cache
-r--r--r-- 1 root root 0 Jan 5 14:47 ndisc_cache
-r--r--r-- 1 root root 0 Jan 5 14:47 rt_cache
> Check throughtput and how packet load is used with 2.4.
> /proc/net/softnet_stat and rtstat.
Will do.
> Also meditate how the comparison can be
> fair wrt taffic patterns. Do smame with 2.6.
Not sure I follow. I can see higher pps/byte through puts on switch port
counters when I run with 2.4 vs 2.6. I'll double check though. I am the
'dev' state so I can swap between V2.4 and V2.6 as necessary to compare
during roughly equivalent times. The only delay is the time between reboots.
> Prepare for a oprofile. I'm lazy and compile the stuff I use into kernel
> if you can get the same result w. nodules it's ok.
Okay.
>
>
> --ro
--
--------------------------------------------------
Jeremy M. Guthrie jeremy.guthrie@xxxxxxxxxx
Senior Network Engineer Phone: 608-298-1061
Berbee Fax: 608-288-3007
5520 Research Park Drive NOC: 608-298-1102
Madison, WI 53711
pgpScBJRobCXB.pgp
Description: PGP signature
|