On Tue, 17 Jan 2006 09:08:06 EST, Clark Kidd wrote:
>
>For many years the MVS operating systems (OS/390 and now z/OS) have
>maintained separate reporting structures for processor time consumed in
>task mode (TCB time) versus processor time consumed while running under an
>authorized subtask (SRB time).  Although these measurements are maintained
>separately in the control blocks and SMF records, we (along with others in
>the industry) have always recommended that these two measurements be
>combined and charged at the same rate for billing purposes.
>
>But there are others who do not agree, and we will still occasionally talk
>to a user who does not charge for SRB CPU time, or who charges a different
>rate for it than regular TCB CPU time.  We would like to see if this is an
>issue for many users.  If you have some type of chargeback system where
>users are billed for CPU usage then please answer the following two
>questions:
>
>1) Do you ignore SRB CPU time and not charge for it as part of your
>accounting algorithm?
>2) Do you charge for SRB CPU time, but use a different rate than for TCB
>CPU time?
>
>If you answered "yes" to either of these two questions, could you please
>send us an email at [EMAIL PROTECTED] and include your reasons
>for doing this?  Thanks in advance for your help.
>
>----------
>
>Clark Kidd
>Watson & Walker, Inc.


a) SRB CPU time can vary from run to run too much, based on what other
things are running concurrent with your job.  YMMV for 'too much'.

b) Can SRBs be dispatched on a zAAP?  If so, the CPU time can vary rather
dramatically on, say, a z890 with degraded general-purpose CPUs where the
SRB is dispatched on a zAAP sometimes and on a CP other times, can't it?
(Why inflict vague stuff like that on a customer's bill?)

--
Tom Schmidt
Madison, WI

----------------------------------------------------------------------
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