T.R | Title | User | Personal Name | Date | Lines |
---|
527.1 | I am perplexed! | WAKEME::ANIL | | Thu Dec 06 1990 18:05 | 83 |
| Hi S�ren,
I am really perplexed as to why you seem to get such odd behavior. I tried two
rules with the same attribute "seconds operating": One with (> 100 ) expression
the other with < 100. Both rules behaved exactly as expected. I have attached
a log of my activity. I would like to request you to do the following:
1. Create another rule with same rule expression but different
rule name.
2. Enable this this rule. Is the behavior as expected?
3. If you still get an error we would like to use your
environment for further debugging.
Thanks,
- Anil
-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=
The bridge Hardware information
-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=
MCC> show BRIDGE {08-00-2B-07-9C-6A} Hardware Type, Seconds Operating
BRIDGE {08-00-2B-07-9C-6A}
AT 6-DEC-1990 17:16:54 Characteristics
Hardware Type = DEBET LAN Bridge 100
BRIDGE {08-00-2B-07-9C-6A}
AT 6-DEC-1990 17:16:56 Counters
Seconds Operating = 3081883 Seconds
-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=
Alarms rule br_100 with > 100
-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=
MCC 0 ALARMS RULE br_100
AT 6-DEC-1990 17:15:53 All Attributes
NAME = br_100
Result of Last Evaluation = True
State = Enabled
Substate = Running
Time of Last Evaluation = 6-DEC-1990 17:15:48.55
Creation Timestamp = 6-DEC-1990 17:15:32.83
Evaluation Error = 0
Evaluation False = 0
Evaluation True = 2
Expression = ( bridge 08-00-2b-07-9c-6a seconds operating > 100, at every 00:00:15 )
Perceived Severity = indeterminate
-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=
Alarms rule br_100 with < 100
NOTE that the rule was evaluated to be False.
-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=
MCC 0 ALARMS RULE br_101
AT 6-DEC-1990 17:15:12 All Attributes
NAME = br_101
Result of Last Evaluation = False
State = Enabled
Substate = Running
Time of Last Evaluation = 6-DEC-1990 17:15:03.40
Creation Timestamp = 6-DEC-1990 17:14:46.93
Evaluation Error = 0
Evaluation False = 2
Evaluation True = 0
Expression = (bridge 08-00-2b-07-9c-6a seconds operating < 100, at every 00:00:15 )
Perceived Severity = indeterminate
MCC>
-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=
Event Notification
-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=
+-----------------------------------------------------------------------------------------+
|!!!!!!!!!!!!!! Alarm, 6-DEC-1990 17:15:49 !!!!!!!!!!!!!! |
|Domain: WAKEME_NS:.foo Severity: Indeterminate |
|Entity: BRIDGE {08-00-2B-07-9C-6A} |
|Rule Name: br_100 |
|Rule Expression: ( bridge 08-00-2b-07-9c-6a seconds operating > 100, at every 00:00:15 ) |
|Rule br_100 has fired |
|Data: BRIDGE {08-00-2B-07-9C-6A} Seconds Operating = 3081814 6-DEC-1990 17:15:48.27 |
+-----------------------------------------------------------------------------------------+
|
527.2 | Some clearification (?) | COPCLU::SORENC | S�ren H Christiansen EIS/ECS - DK | Fri Dec 07 1990 04:19 | 13 |
| I have tried the things you recommended.
A different rule name does not change things. A condition against a
different attribute (user bytes sent) works OK.
Now when modifying the constant (100) I have found out that the
behavior is unexpected for constant values from 100 to 999, where
as 99 and 1000 work fine. (The attribute value was is the range of
700000 seconds).
Can I provide you more valuable information ?
S�ren
|
527.3 | Could it be a counter48 problem? | WAKEME::ANIL | | Fri Dec 07 1990 07:26 | 16 |
| I can see a pattern now! I bet its a data type problem. Seconds operating
is a counter48 data type with length of 8 bytes. This by itself should
not be a problem, but the math does get tricky. Also I will investigate
if all the 8 bytes are comming back from Bridge AM.
I tried the folowing just to verify!
show bridge xyz seconds operating, with (seconds operating > 100000000)
The actual value of the attribute was 2423122. So I did not expect
to see the result. Guess what I saw 2423122 printed on my screen.
I will report what the exact problem is soon.
Thanks S�ren, for pointing out a rather buzzar behaviour! :(
- Anil Navkal
|
527.4 | I need to see your problem in your environment! | WAKEME::ANIL | | Fri Dec 07 1990 17:15 | 26 |
| Hi S�ren,
Off and on I have spent good part of my day to see how using the
our evaluation algorithm, we could have possibly made an error. I
have looked at the code, used the debuger to deposit your values,
ran 50 rules with expression same as yours at different polling
intervals. I regret to note that I wasn't able to encounter the
same problem you reported.
Giving you a run down on my activity to isolate your problem
makes me feel good but does not solve your problem! :( It just
tells me that I just cannot reproduce it!
The only thing I can now think of trying is, to get into your
environment to actually see the behavior, and if needed copy a
debug image of Alarms FM at your end. Is this okay with you? If so
send me mail and we will workout some thing. I need an account
to log into your system and disk space to copy ~ 3000 blocks.
If any of you have experienced similar problem please report it
here ASAP.
Regards,
Anil Navkal
Alarms Project Leader
|
527.5 | Its history now! | WAKEME::ANIL | | Thu Dec 20 1990 09:01 | 14 |
| After a rather prolonged and painful investigation a bug was
uncovered in the routine that converts the ascii string in the a
counter[48|64] number. QAR #116 in MCC_INTERNAL data base will now
track this bug.
S�ren, thank you for letting us know about this odd behavior. The
bug has now been fixed.
But for the efforts by Keith and Pete Ditmars, isolating the bug
would have been even more painful! Good team work guys.
- Anil
|