[ale] Experiment

Joseph A Knapka jknapka at earthlink.net
Mon Nov 4 13:59:29 EST 2002


I just performed a little experiment:

Machine A: P2/233, 128M, running a simple Java RMI server.

Machine B: Linux 2.4.18, K6-2/500, 128M 256M swap. I ran a script
that started 200 RMI clients, each of which made 1000 requests of
the server. Five of them failed to start due, apparently, to memory
inavailablity; the remainder all started and ran to completion,
which took about 5 minutes. During this time, the machine
remained somewhat responsive - I could log in, run ps, and
grep the ps output to see how many clients were running.
This machine was also running a Zope instance and a MySQL
server, both of which survived without injury (though I
didn't try to use either of them while the herd was
thundering by).

Machine C: P4/1.6GHz, 256M, 384M swap Windows XP Home, patched
to most recent Windows Update level last week. Ran the same RMI-
client-herd script. Machine locked up completely within ten
seconds, and stopped delivering requests to the RMI server.
Not even the mouse pointer worked, nor would Ctl-Alt-Del
bring up the system dialog. It took a power cycle to regain
control.

And MicroSh*t calls their products "reliable"?

Cheers,

-- Joe


---
This message has been sent through the ALE general discussion list.
See http://www.ale.org/mailing-lists.shtml for more info. Problems should be 
sent to listmaster at ale dot org.






More information about the Ale mailing list