Lucent Technologies DEFINITY Performance Management Release 2.0 User Guide

							DEFINITY Performance Management User Guide, Release 2.0, 585-229-808Issue 1
November 1998
Page 201
10
10Maintenance and Error RecoveryChapter Contents•Introduction
202
•Error Recovery
203
•System Notification of Errors
205
•Process Trace
206
•Event Types Exception and Error
207
•Message Type SWERR
208 
						

							DEFINITY Performance Management User Guide, Release 2.0, 585-229-808
10Maintenance and Error RecoveryIntroduction
Issue 1
November 1998
Page 202
Introduction
This chapter describes various aspects of maintaining Performance Management. 
It also describes how the system recovers from various errors. Topics included in 
this chapter are:
• What Performance Management does upon initialization
• Errors and how the system handles them
• Trace file description
• Event type description 
						

							DEFINITY Performance Management User Guide, Release 2.0, 585-229-808
10Maintenance and Error RecoveryError Recovery
Issue 1
November 1998
Page 203
Error Recovery
Performance Management handles errors in the following manner:Type of Error Description How System HandlesSNMP errors Errors returned to the 
application software 
from the lowest 
SNMP software. 
Usually result in data 
collection failure.• Informs user of error by alert in Data 
Collection node in left pane.
• Logs the error in the event log.
• Stops requesting data
• Attempts to collect data again at the 
next scheduled collection time.
SNMP failures SNMP response 
indicating an error 
condition at the Proxy 
Agent or the network.• Informs user of failure by an alert in 
the Data Collection node.
• Logs the failure in the event log.
• Stops requesting SNMP data from 
the node that provided the failure 
response.
• Continues to collect data from other 
managed nodes
• Attempts to collect data from the 
affected managed node at the next 
scheduled collection time. 
						

							DEFINITY Performance Management User Guide, Release 2.0, 585-229-808
10Maintenance and Error RecoveryError Recovery
Issue 1
November 1998
Page 204
SNMP No 
ResponseProblem with the 
network or Proxy 
Agent.• Retries the request. If retries fail, 
displays an alert in the Data 
Collection node.
• Logs the error in the event log.
• Posts an alert in the Data Collection 
node.
• Stops requesting data from the 
failed managed node, but continues 
to request data from the other 
managed nodes as scheduled.
• Attempts to request data from failed 
managed node at next scheduled 
collection time.
Inconsistent 
DataFor example, a trunk 
group data with no 
corresponding 
configuration data.• Stores the data normally.
• Writes a message to the trace file.Type of Error Description How System Handles 
						

							DEFINITY Performance Management User Guide, Release 2.0, 585-229-808
10Maintenance and Error RecoverySystem Notification of Errors
Issue 1
November 1998
Page 205
System Notification of Errors
Performance Management provides information for errors that can prevent the 
system from performing its task. Errors can display in the following ways:
• Pop-up error messages
• Alerts shown in the left pane
Error and 
Exception Log
All performance exceptions and selected errors are recorded in the database in 
an event log. You can view the events by viewing the various event log panes, 
which generate alerts that have an associated severity level based on how the 
alerting parameters are defined. The alerts appear in the left pane and are 
associated with the DEFINITY component (managed node or performance report) 
that generated the event.
SNMP problems generate a data collection failure event. A scheduled report that 
is not generated also generates an event. Exception events are performance 
exceptions that are outside threshold limits as defined for the system.
Performance Management periodically checks the event log and deletes records 
whose start time exceeds the defined time that event records are kept. 
						

							DEFINITY Performance Management User Guide, Release 2.0, 585-229-808
10Maintenance and Error RecoveryProcess Trace
Issue 1
November 1998
Page 206
Process TraceDescription
Each pprocess writes a message to a common trace file. The message has:
•Time stamp
• Process name
• User name
• Managed node name (or “DPM” if message is not related to a managed node)
• Text description that varies with the message type
About LOG and 
TRACE
Message types LOG and TRACE are normal application progress messages. 
LOG messages are always logged, and correspond to trace level zero.
TRACE messages are logged if you raise the trace level.
TRACE 
Description
A TRACE message contains trace level, thread name if applicable, class, and 
method. The system has an indeterminate number of TRACE events, but only a 
few fixed LOG events. 
						

							DEFINITY Performance Management User Guide, Release 2.0, 585-229-808
10Maintenance and Error RecoveryEvent Types Exception and Error
Issue 1
November 1998
Page 207
Event Types Exception and Error
The start and end of an exception or error alert is logged in the trace file. The 
events are described in the following table:Exception Event DescriptionProcessor occupancy Call processing occupancy, occupancy threshold
Trunk group GOS Trunk group number, exception criteria, size, 
recommended size
Scheduled report failure Report name, error code, error description, if 
available
Scheduled report 
truncatedReport name, record limit, and total records, if 
available
Scheduled report data 
missingReport name, data type
SNMP no response Switch name, timeout interval, number of retries
SNMP failure Error code and error string, if available
Database storage failure Error code and error string, if available
Inconsistent data Switch name, data type 
						

							DEFINITY Performance Management User Guide, Release 2.0, 585-229-808
10Maintenance and Error RecoveryMessage Type SWERR
Issue 1
November 1998
Page 208
Message Type SWERR
These are internal software errors that do not cause an alert, though they may 
cause other failures that will cause alerts.SWERR Event DescriptionInternal SNMP errors Error code and error 
string, if available
Abnormal process 
terminationProcess name 
						

							DEFINITY Performance Management User Guide, Release 2.0, 585-229-808
10Maintenance and Error RecoveryDelete Managed Node from Database
Issue 1
November 1998
Page 209
Delete Managed Node from Database
When a managed node is permanently removed from the Proxy Agent, its data 
remains in the database. Since the data for the managed node is not necessary, 
you can delete the data from the database by following this procedure.
Procedure
To remove the data for a managed node from the database, do the following:1Shut down background processes by typing the following at the UNIX prompt 
and pressing 
Return
:
ProcStartup -K
2Type the following and press 
Return
:
RemoveDEFResult: 
Displays any managed nodes that have been removed from the 
Proxy Agent.
3Type the command in step 3 followed by each managed node to be deleted. 
To delete multiple nodes, separate each managed node by a space. For 
example:RemoveDEF Definity1 Definity2Result: 
The data from the specified managed nodes are removed from the 
database. 
						

							DEFINITY Performance Management User Guide, Release 2.0, 585-229-808 Issue 1
November 1998
Page 210
10Index
Symbols
$APP_DEF/OVw file, to assign custom color
50
Numerics
3-dimensional, in chart setup
160
A
active events
185
aggregator
in report format
160
in sample chart
161
Alarm Forwarding Status field
56
alarm notification, to set up
73
alarm states
65
alarm, methods of notification
72
alarms
for Proxy Agent
64
Proxy Agent
65
alert icons
188
alert indicator
86
,95
alert indicators, on event log
181
alert levels, icons
95
alert levels, table
188
alert notification, methods for NMS
71
alert notification, to set up
73
Alerting
76
alerting parameters
set global
190
set processor occupancy
195
set trunk group
198