The Question is:
I am responsible for a clients vax 6640 and before I took over someone was
sending a report to the account reps showing a report of disk usage(I have
know problem figuring out how to collect this data over a month's period) and
a VUPS's usage in "%". I h
ave NO clue how to collect this later piece. I am familiar with monitor, but
how does this relate to VUP's usage. This report was sent monthly. Sorry i
haven't any info on algorythms used. ANY advice would be greatly appreciated.
The Answer is :
The OpenVMS Wizard can only assume this is a scaled conversion of
CPU usage, and that it has little or nothing to do with the VUPs
measurement. Or that the calculation is scaled based on the (old
and no longer used) VUPS rating of the particular processor.
The original VUPs rating of the VAX 6000 model 600 series is roughly
seven (7) VUPs per processor to a maximum of thirty-six (36) VUPs
for the VAX 6000 model 660 series -- the maximum value assumes that
all six processors in the model 660 are kept busy. No single-stream
process will receive more than seven VUPs performance. As with most
(all?) other multiprocessor system configurations, the maximum system
performance is achieved via extensive parallelism.
For purposes of comparison, the VAXstation 4000 model 96 series is
rated at roughly forty-six (46) VUPs performance (single-stream), and
the MicroVAX 3100 models 90 and 95 rated at twenty-four (24) and
thirty-two (32) VUPs (single-stream).
Direct access to the mechanisms underlying MONITOR is available via
the undocumented and unsupported exe$getspi API. For details and for
program examples of calling exe$getspi, please see the *SPI* tools in
the OpenVMS Freeware directory SRH_EXAMPLES. OpenVMS V7.3 and later
are expected to have documented and supported access to this system
performance information via the sys$getrmi system service and related
data collection mechanisms.