Please excuse my indecision. The "refresh rate" referred to was the
refresh frequency of the online display, not the sample frequency

HTH,
<snip>
The "refresh rate" for any ISV monitor may refer to any number of things
but it is unlikely to refer to a global sampling interval if we are
talking about numbers in multiples of 5 or 10 seconds. 

I imagine that most monitors sample the various objects within z/OS at
out-of-box "sensible" intervals depending on the object concerned and
they may (or may not) allow the installation to change these intervals.
For example, you would probably want to sample your address spaces more
frequently that your DASD volumes or XCF structures. Sampling address
spaces at intervals greater than a second or two would probably result
in unacceptable increased granularity of the results.  

I suspect the "refresh rate" in this case is how often the display goes
back to the performance data-store and asks for a refresh of the data -
this will be independent of the sampling rates of the underlying
performance data. 
</snip>

Rob Scott
<snip>
<snip>
Something else comes to mind. I recall measuring these some monitors and
seeing CPU consumption affective. The REFRESH RATE may also force some
sorting, XMS and other stuff to occur on every 20 seconds. Measure the
delta changes on CPU TIME when you adjust the refresh rates. CPU
consumption will should down within the address space with longer
refresh rates </snip> 

I once had an anally retentive manager who complained about the CPU
consumption of our monitor. In order to appease him We change the
refresh rate on the monitor from 5 sec to 30 sec. No observable
difference in performance at any level was observed. After the
experiment was concluded, we returned the monitor refresh rate to 5 sec.

HTH,
</snip>

----------------------------------------------------------------------
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [EMAIL PROTECTED] with the message: GET IBM-MAIN INFO
Search the archives at http://bama.ua.edu/archives/ibm-main.html

Reply via email to