Oh YEAH... Got you with *that* subject line, now didn't I?
<heh heh heh... *ahem* ANYway...>
What SCOPE metric(s) would you use to present the impact of a user's
session(s) on a system? That is, TOP_CPU (Process) and CPU (Process)
show overall % CPU usage and % CPU usage within an interval
respectively, but they're more indicative of what a session *gets*, not
what it *takes* (or "needs"). Listed percentages are completely
relative to what else is happening on the box, so they don't necessarily
tell you what some specific process/thingie-ma-bob requires. (As
always, "It depends..." <grin>)
So if you had to present a digestible graph to management on what
resources (esp. CPU) a particular session takes whenever logging on, how
would you do it? What sort of basic (read: "Simple") baseline metric
or calculation can you use to compare utilization metrics against?
Would TOP_CPU (Process) divided by PROCESSOR1_BUSY (Global) work as a
thumb-rule?
Perhaps CPU_SECONDS (Process) minused from (or divided by)
CPU_SESSION_SECONDS (Global)?
How 'bout taking a before and after "picture" -- but then what metrics
get measured? All?
(How would you summarize the pictures -- especially if other processes
in the shots aren't *quite* the same?)
I really need some sort of actual measure of this session's activity in
order to prove/disprove why we wouldn't want many more of them doing the
same thing(s).
("Automated PC-based Scripts: Who Needs Batch Jobs Anymore?" <*sigh*>)
Truly TIA,
Curtis
-----------------------------------------------------
Confidentiality Notice: This e-mail transmission
may contain confidential or legally privileged
information that is intended only for the individual
or entity named in the e-mail address. If you are not
the intended recipient, you are hereby notified that
any disclosure, copying, distribution, or reliance
upon the contents of this e-mail is strictly prohibited.
If you have received this e-mail transmission in error,
please reply to the sender, so that we can arrange
for proper delivery, and then please delete the message
from your inbox. Thank you.
* To join/leave the list, search archives, change list settings, *
* etc., please visit http://raven.utc.edu/archives/hp3000-l.html *
|