variablilty in process run times

Jim Nance (jlnance@avanticorp.com)
Wed, 6 Jan 1999 10:48:54 -0500

Hello All,
Yesterday I made a change to some code and I wanted to see if it was
faster or slower than the original. Since I had 4 identical Linux boxes
I ran the old code on penguin1 and penguin3 and the old code on penguin2
and penguin4. The program keeps track of how much CPU it used and here
is what I get from the log files:

penguin1: Total check time = 3:47:10 User=11996.59 Sys=1054.27 Mem=209.109
penguin2: Total check time = 4:46:52 User=14361.63 Sys=1236.27 Mem=209.109
penguin3: Total check time = 3:54:27 User=11991.15 Sys=1061.73 Mem=209.109
penguin4: Total check time = 3:54:43 User=11955.24 Sys=1062.43 Mem=209.109

I was quite supprised by the amount of variablilty in the runtimes since
nothing else was running on the machines. Even if I assume that some unknown
event happened on penguin2, the 7 minute difference between penguin1 and
penguin3 still supprises me. I usually run these tests on a Digital Unix
system and I get runtimes that vary by less than a minute. Should I expect
this much variance?

Configuration:
Linux 2.2.0-pre4 Kernel on Red Hat 5.2 system.
333MHz P-II with 512M ram
ncr53c875-0-<0,*>: FAST-20 WIDE SCSI 40.0 MB/s (50 ns, offset 15)
Vendor: COMPAQ Model: WDE4360W Rev: 1.52
Type: Direct-Access ANSI SCSI revision: 02

Thanks,

Jim

-- 
----------------------------------------------------------------------------
Jim Nance                                                 Avant! Corporation
(919) 941-6655    Do you have sweet iced tea?       jim_nance@avanticorp.com
                  No, but there's sugar on the table.


-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu
Please read the FAQ at http://www.tux.org/lkml/