| Title: | VAX and Alpha VMS |
| Notice: | This is a new VMSnotes, please read note 2.1 |
| Moderator: | VAXAXP::BERNARDO |
| Created: | Wed Jan 22 1997 |
| Last Modified: | Fri Jun 06 1997 |
| Last Successful Update: | Fri Jun 06 1997 |
| Number of topics: | 703 |
| Total number of notes: | 3722 |
I just received the following question from a customer and don't have
an answer. Does anyone know why my customer is seeing this behavior?
I have a command procedure that I run at 10 minute intervals to
monitor the performance of a process. In this monitor sequence
is are the following two lines.
$ biolm = f$getjpi(pid,"biolm")
$ bioused = biolm - f$getjpi(pid,"biocnt")
For some reason that I don't understand, this sequence will some
times return a negative value for "bioused". To me, this implies
that either "biolm" is being lowered at some point between these
two commands or "biocnt" is going above "biolm". I'm not
changing "biolm" which only leaves "biocnt" going above "biolm".
Is this possible? If so, how, since I thought trying to go above
a limit would cause the error message, "you have exhausted your quota".
Can you shed any light on this?
Any help would be appreciated
Barbara Youngwirth
| T.R | Title | User | Personal Name | Date | Lines |
|---|---|---|---|---|---|
| 365.1 | WAG | AUSS::GARSON | DECcharity Program Office | Sun Mar 23 1997 22:00 | 30 |
re .0
Looks strange.
> In this monitor sequence is are the following two lines.
Are the two lines consecutive? E.g. are we sure that "pid" does not
change meantime, for example, by programming error.
The obvious thing to do would be to check for bioused .lt. 0 in the
code and write out the actual values of biolm and biocnt (which one
would need to capture explicitly).
> I'm not changing "biolm"
I don't believe there is a supported means of changing biolm (dynamically)
anyway.
> which only leaves "biocnt" going above "biolm". Is this possible? If so,
> how, since I thought trying to go above a limit would cause the error
> message, "you have exhausted your quota".
BIOCNT above BIOLM would mean that the amount free was above the quota
so one would not expect any "quota exhaustion" message. Quota
exhaustion corresponds to the CNT getting down to zero.
The only long shot that I can offer is that due to some condition the
information cannot be obtained from the process (as $GETJPI is known to
do if the target process is in certain scheduling states) and that this
is not handled correctly. Maybe LM ends up as 0 or something.
| |||||