T.R | Title | User | Personal Name | Date | Lines |
---|
2585.1 | the Rule looks fine | NANOVX::ROBERTS | Keith Roberts - DECmcc Toolkit Team | Wed Mar 18 1992 07:17 | 22 |
| RE: .0
Ian,
>> Error Condition = "%MCC-E-INV_IN_P, invalid argument or
>> attribute , The rule has been disabled"
>>
>> Expression = (OCCURS (Node LOCAL_NS:.reo.reotn3 HDLC Link * Link Down))
Your Rule Expression looks good. My first guess would be that the DECnet
Phase-5 AM returned the INV_IN_P error to Alarms. The part of the Rule
Expression which is passed to the DNA-5 AM through the In-P argument is
the Event Code; in this case, Link Down.
I believe that there was recently a QAR on Alarms that stated the Event Code
was not properly constructed by the OCCURS function.
(Q) Does this error occur every time?
(Q) Could you try replacing 'Link Down' with 'Any Event'
/keith
|
2585.2 | more info | RDGENG::PRATT | | Wed Mar 18 1992 08:41 | 18 |
| Keith
>> (Q) Does this error occur every time?
Yes - I have also tried the event Link Up with the same results.
>> (Q) Could you try replacing 'Link Down' with 'Any Event'
This works and I can create the rule okay, but I cannot get it to fire. I
already have the phaseV events sinking to a window on my workstation, and I have
enabled the dna5_am sink, and the process /usr/mcc/mmexe/mcc_dna5_evl is running
Is there anything else I should do? Do I have to force the events to sink into
MCC - if so how?
Cheers
Ian
|
2585.3 | The Occurs function does a GETEVENT | NANOVX::ROBERTS | Keith Roberts - DECmcc Toolkit Team | Wed Mar 18 1992 11:11 | 29 |
| RE: .2
>>> (Q) Could you try replacing 'Link Down' with 'Any Event'
>
> This works and I can create the rule okay, but I cannot get it to fire. I
> already have the phaseV events sinking to a window on my workstation, and I
> have enabled the dna5_am sink, and the process /usr/mcc/mmexe/mcc_dna5_evl
> is running
Ian,
The Alarms OCCURS function uses the GETEVENT directive. Try doing
the same command alarms would - to make sure the DNA-5 evens are
working correctly within DECmcc ...
getevent Node LOCAL_NS:.reo.reotn3 HDLC Link * any event
Also have the same Alarm Rule running before you issue this command.
If you get a good response from the getevent directive, check the rule
status and counter attributes ... the 'evaluation true' counter should
have incremented.
If you get no response from the getevent directive, then there is a
problem getting DNA-5 event into DECmcc.
Let me know how this goes.
/keith
|
2585.4 | Works now | RDGENG::PRATT | | Wed Mar 18 1992 11:50 | 19 |
| Keith,
I've managed to get the 'Any Event' rule to fire now - in the release notes you
have to set up an outbound stream to object 99 - I was using the default object
82.
>> getevent Node LOCAL_NS:.reo.reotn3 HDLC Link * any event
This command 'hangs' until I generate an event manually from the DECnis using
ncl - ie disable node .reo.reotn3 HDLC Link *
and then it returned with a link down event - is this normal operation, as I'm
sure that when I tried this yesterday without the event stream configured I got
a list of all possible events almost immediately.
Is the Expression = (OCCURS (Node LOCAL_NS:.reo.reotn3 HDLC Link * Link Down))
a definite bug as it still fails?
Ian
|
2585.5 | Looks like Alarms has a problem with Event Ids | NANOVX::ROBERTS | Keith Roberts - DECmcc Toolkit Team | Thu Mar 19 1992 09:45 | 32 |
| RE: .4
Ian,
> I've managed to get the 'Any Event' rule to fire now - in the release notes you
> have to set up an outbound stream to object 99 - I was using the default object
> 82.
>
> >> getevent Node LOCAL_NS:.reo.reotn3 HDLC Link * any event
>
> This command 'hangs' until I generate an event manually from the DECnis using
> ncl - ie disable node .reo.reotn3 HDLC Link *
Great!
> and then it returned with a link down event - is this normal operation, as I'm
> sure that when I tried this yesterday without the event stream configured I got
> a list of all possible events almost immediately.
Someone from DNA5 AM team will have to answer this one.
> Is the Expression = (OCCURS (Node LOCAL_NS:.reo.reotn3 HDLC Link * Link Down))
> a definite bug as it still fails?
Yes, I'd have to say so. You still get the INV_IN_P error condition when
you do show rule status - Right?
I believe there is a QAR on this problem - and will double check. Alarms
not creating the In-P buffer for the OCCURS function will be fixed for
v1.2 Field Test Update.
/keith
|
2585.6 | Same error | RDGENG::PRATT | | Thu Mar 19 1992 10:23 | 21 |
| >> Yes, I'd have to say so. You still get the INV_IN_P error condition when
>> you do show rule status - Right?
Correct
>> This command 'hangs' until I generate an event manually from the DECnis using
>> ncl - ie disable node .reo.reotn3 HDLC Link *
>> Great!
>> and then it returned with a link down event - is this normal operation, as I'm
>> sure that when I tried this yesterday without the event stream configured I got
>> a list of all possible events almost immediately.
>> Someone from DNA5 AM team will have to answer this one.
Not a problem as I've now realised what I was doing different.
Thanks for your help
Ian
|
2585.7 | See note 2660.5 for an answer | MOLAR::ROBERTS | Keith Roberts - DECmcc Toolkit Team | Thu Apr 02 1992 08:42 | 0
|