[Info-vax] HIBertanting DECthreads process accrues CPU time.

[VAXman- at SendSpamHere.ORG]

In article <422e7e27-0a8a-45a6-a721-b6cf3addedb6 at u17g2000yqe.googlegroups.com>, John Wallace <johnwallace4 at yahoo.co.uk> writes:
>
>Is it really accruing CPU time at an alarming rate? Or is the usually
>reliable mechanism for reporting CPU time falling victim to odd
>behaviour in this case, and causing unnecessary alarm?

Reality or spectre, seeing an otherwise *IDLE* process accruing CPU time
is not acceptable to me and is likely to cause alarm with customers.

 Elapsed CPU time:          0 00:12:51.03
 Connect time:              0 14:06:07.99

 Elapsed CPU time:          0 00:12:53.25  ~2.2 secs
 Connect time:              0 14:09:57.21  ~4 mins

 Elapsed CPU time:          0 00:12:55.66  ~2.4 secs
 Connect time:              0 14:14:16.74  ~4mins

That's quite a bit of CPU time on an Itanium in 4 mins!  When the process
is started and doesn't tailspin, asking to to do a significant amount of
work (process 100K records for example) will not see it accrue 2 seconds!

There is the primary thread (the $CREATE process thread running at normal
thread priority (11)) and two other threads (running a the lowest priority
(8)).   They're sitting on the $HIBER and have not done any work.  There's
a little debug marker I put in each thread about the $HIBER that increment
a variable before and after the call.  The afters are 0 and the befores 1.
So my threads ARE *IDLE*.

-- 
VAXman- A Bored Certified VMS Kernel Mode Hacker    VAXman(at)TMESIS(dot)ORG

All your spirit rack abuses, come to haunt you back by day.
All your Byzantine excuses, given time, given you away.