VMS::Monitor - Access system performace information on OpenVMS systems
use VMS::Monitor; $hashref = VMS::Monitor::all_monitor_info(); $one_piece = VMS::Monitor::one_monitor_piece($Thing_to_get); @InfoNames = VMS::Monitor::monitor_info_names(); (Well, not yet)
Retrieve performace info via the $GETSPI system call.
Important stuff that I might otherwise forget (yes, it is sort of stream of consciousness documentation. Be afraid, be very afraid...):
unless otherwise noted, all the return values are integers
Most of the data returned is a count, rather than a delta value. So, for example, when you retrieve the FAULTS item, it's the total number of faults since reboot. If you want to do any sort of monitoring, you'll need to take multiple samples and do the math yourself.
Each entry in the array represents a single cluster member (including HSJs and other cluster storage). The entry is a reference to a hash, which has a bunch of key/value pairs. (Whod've though?) Currently the keys are DGDISCARD DGRCVD KBYTMAPD KBYTREQD KBYTSENT MSGRCVD MSGSENT NODENAME QBDT_CNT QCR_CNT REQDATS SNDDATS, and pretty much correspond to the names MONITOR uses.
note that at some point this might change from returning an arrayref to returning the actual array, but I don't know when. (Or if, really) I'm up for suggestions.
Like the SCS class, DISKS returns a reference to an array of hashrefs, one ref per disk. Each of those hashes has the keys NODENAME, VOLNAME, DEVNAME, ALLOCLASS, FLAGS, OPTCNT, QCOUNT, and UNITNUM, which correspond to the data that MONITOR returns.
Note that VOLNAME has trailing blanks for some reason. (No, I don't know why. Probably backwards compatibility or something like that) They may get stripped off at some point before this goes final, though.
Dan Sugalski <dan@sidhe.org> Maintained by Craig A. Berry <craigberry@mac.com>
perl(1), OpenVMS System Management Utilities Reference Manual: M-Z, Appendix A
