| T.R | Title | User | Personal
Name | Date | Lines |
|---|
| 3162.1 | User visible syntax is very important | NANOVX::ROBERTS | Keith Roberts - DECmcc Toolkit Team | Wed Jun 10 1992 14:51 | 31 |
| RE: .0
> Are there any plans to allow the user to define rules that fire
> when an event does not happen within a duration (timeout) ?
> something like: NOT_OCCURS( ENTITY X EVENT event_name , duration)
Manuel,
I have been thinking more about this since our phone conversation.
I was thinking of using the existing OCCURS function:
occurs( <event>, <count>, <duration> )
but specify 0 as the count. If the Event occurs zero times in a
certain lengh of time ... let me know.
Of course, for such an expression there is still a scope of interest
time (forever is the default) which defines how long you want this
situation monitored.
Now .. as far as the coding efforts .. This will be interesting. I imagine
that a 'Thread Terminator' would be used to time-out the request for the
event after the <duration> had expired. If the counter was zero, then a
rule-fired event would be generated and the process would continue. If
the counter was non-zero no event would be generated and the (internal)
counter reset for the next evaluation.
I think this is a *very* good suggestion - and should be added to the list
of Alarms features for after v1.2
/keith
|
| 3162.2 | lost events ? | ULYSSE::MORALES | Reddite quae sunt Caesaris Caesari | Thu Jun 11 1992 11:46 | 19 |
|
Keith,
according to the use I see for this feature, such an OCCURS rule that will
fire if 0 (ZERO) event is seen for a given duration will satisfy my
requirements (but others ??)
If MCC is supposed to receive an event every 10 minutes, after a duration
of 10 minutes a count of zero event will generate the rule-fired event.
If I want to allow for lost events without firing the rule, a n*10 minutes
duration will allow for n-1 lost events.
But what if I want to allow to loose only 1 event among n ?
By creating a NOT_OCCURS rule we could get more flexibility.
not_occurs(entity x event,count,duration)
Manuel.
|
| 3162.3 | I see now | NANOVX::ROBERTS | Keith Roberts - DECmcc Toolkit Team | Thu Jun 11 1992 12:21 | 7 |
| Manuel,
I get it now 8)
Yes .. a new function would be better for this functionality.
/keith
|