Conference azur::mcc

3929.0. "managed object field too long" by CSC32::W_MCGAW () Tue Oct 20 1992 13:39

Subj:   Problems with DECMCC EMS V2.2 alarms/exception Data Files


I have a customer that has been running DECMCC EMS V2.2 for about 10 
days and is having problems with some of the alarm/exception data files 
that are being created.

The record that is created for MANAGED_OBJECT sometimes is being generated
with over 1000 characters (most of which are blank).  The first part of the
record is ok, then there are blanks and the last character is a digit like
1 or 2.

This creates a serious problem as a record of this lenght can be read by
a DCL command procedure, but it cannot be processed by DCL statements
because of DCL command buffer overflow.

It seems that this only occurs for some rules which are written against
bridges.

RULE: Domain EPA:.MCC.RTPNC Rule BRIDGE_MACHIP_BEAUIP_LINE_2_ERRORS 
MANAGED_OBJECT: Bridge EPA:.rtpnc.bridge.MACHIP_BEAUIP LINE 2                  
										
	.
	.
	.
lots of spaces removed
	.
	.
	.
2
DESCRIPTION: High error rate.
CATEGORY: Backbone Sick
EXPRESSION: (BRIDGE rtpnc.bridge.MACHIP_BEAUIP line 2 error percent > 1.0, -
at every 0:15:11, for start -0:00 duration 0:6:11,via port = "eza0:")
TIME: 19-OCT-1992 11:04:51.28
EVIDENCE: Communication with the target has been interrupted
EVENT_ARGUMENTS: 
PARAMETER: MAIL=@MCC_ALARMS:BRIDGE_ALERTS.DIS;LOG=MCC_LOG:BRIDGE_ERRORS.LOG
DOMAIN: Domain EPA:.MCC.RTPNC 
SEVERITY: Indeterminate 


Notice that the expression includes a ,via port qualifier.  The customer was
getting strange errors until he removed the qualifier from the expression and
put it into the enable command.



Any ideas or suggestions?

Walt