Conference azur::mcc

6114.0. "Thousands of MCC_ALARMS_nnnnnnnn.DAT files ?" by KAOFS::CARSWELL () Wed Sep 07 1994 11:20

	We have encountered a problem, that 'appears' to be related to MCC,
and the symptoms are extremely strange. 

	I have already logged a call with the CSC, but in the interests of
not having to restart MCC unless ABSOLUTELY necessary I thought I'd also ask
if anyone here has seen this problem.

	About 4 days ago, we started accumulating MCC_ALARMS_nnnnnnnn.DAT files
in the SYSMGR directory.  These files  are being created at a varying rate of 
between 4000-10000 per day, and the strange thing is that they are date-stamped
as far back as April of this year?   The information contained in the Alarm
files is potentially valid alarm information from back when the real alarm
fired, but I have absolutely no idea, why the files are being created now, AND
WHY are they not getting deleted. 

	Here are just a couple of the files in question, and as you will
notice, the dates are old.  I KNOW that all of these files were created
yesterday (Sep 6,1994) and assume it was around 10:00. (Notice how the
actual creation time is sequential, but the datestamp is NOT!?)    

MCC_ALARMS_DATA_10000214.DAT;1     2   9-JUL-1994 10:00:02.20
MCC_ALARMS_DATA_10002374.DAT;1     2  12-MAY-1994 10:00:23.93
MCC_ALARMS_DATA_10002765.DAT;1     2   8-JUL-1994 10:00:28.43
MCC_ALARMS_DATA_10002768.DAT;1     2  20-JUN-1994 10:00:27.94
MCC_ALARMS_DATA_10002776.DAT;1     2   2-JUN-1994 10:00:28.08
MCC_ALARMS_DATA_10003435.DAT;1     2  29-JUN-1994 10:00:34.88


And  there is valid data inside these files.
-------------------------------------
| $TYPE MCC_ALARMS_DATA_10000214.DAT;1 |
-------------------------------------
RULE: Domain LOCAL_NS:.cpc Rule ADJACENCY_UP
MANAGED_OBJECT: Node4 LOCAL_NS:.dna_node.cvnx Circuit ETHERNET Adjacent 
                Node CVN80
DESCRIPTION: UP
CATEGORY: ADJACENCY
EXPRESSION: (OCCURS(NODE4 * CIRCUIT * ADJACENT NODE * ADJACENCY UP))
TIME: 9-JUL-1994 10:00:02.14
EVIDENCE: The last event detected: Node4 LOCAL_NS:.dna_node.cvnx Circuit 
          ETHERNET Adjacent Node CVN80 Adjacency up   9-JUL-1994 09:59:29.13
EVENT_ARGUMENTS: Event: Adjacency up  Adjacency Up   Adjacent Node Address = 45.10
PARAMETER: ALERT
DOMAIN: Domain LOCAL_NS:.cpc
SEVERITY:     Critical


	We have VMS V5.5-2, MCC V1.3, we use the MCC_DNA4_EVL process, to 
collect phase IV events, the MCC-EVC-SINK to collect non-MCC type stuff, 
the IP-POLLER process, and the rules in the context of a DETMCC detached
process.  This entire setup has been in use for over a year, and NOTHING
was changed in the setup recently.    AND EXCEPT FOR THIS ANOMALLY MCC 
IS WORKING JUST FINE, IN EVERY RESPECT!


	Any help/suggestions that anyone may provide would be appreciated.
This is a Production System(as usual), and we do not wish to reboot, if
we don't have to.  

	Regards,
		Gord C.

	Just in case anyone encounters this one..... 

	We opted to reboot the system, and this strange problem disappeared.
There were no ;32768  files around(.LOG or otherwise) although a good 
suggestion for sure.  Thanks.


	Although a restart of MCC 'may' have cleared the problem, we opted
for the reboot since we have very limited  allowable downtime.  Killing
the fly with a shotgun(the reboot) may be a bit excessive, but the result is
usually pretty predicitable, and as illustrated here, provided the desired
result!

	Regards,
		Gord C.