Chapter 8: Performance Management

Table Of Contents

Performance Management

8.1  Introduction

8.1.1  Definitions

8.1.2  Present G.826 PM data

8.1.3  View Counters

8.1.4  Criteria for Counting Valid-data

8.2  Manage RMON

8.2.1  AboUt Rmon Measurements In Cisco Network Elements

8.2.2  RMON Overview

8.2.3  Create RMON Event Monitor

8.2.4  Configure an RMON Event Monitor

8.2.5  Create RMON History Monitor(s)

8.3  View RMON Data

8.3.1  View Statistical Data

8.3.2  View logged events

Performance Management

8.1  Introduction

This chapter describes the presentation of Performance Management (PM) data as they occur on the network element. The PM data available on the network element:

•G.826 performance data for the SDH paths and section termination points.

–G.826 on Multiplex Section (MS), Regeneration Section (RS), Virtual Containers (VC).

–Non-intrusive monitors on AU's and TUs cross- connected through the network element.

•Values of the various counters.

•RMON.

PM data can be monitored from the Management Tree. You can read the registered PM data on the network element, get it presented in performance management data tables. The file can be read, copied and/or edited in any tool, for instance MS Excel. It is possible to clear all PM data on the network element, see "Logs (Alarm Logs, Performance Data Logs)" on page -32.

8.1.1  Definitions

According to G.826 PM data, the following definitions are used:

•Errored second (ES)

–A one second period with one or more errored blocks or at least one defect.

•Severely errored second (SES)

–A one second period which contains >= 30% errored blocks or at least one defect

•Background block error (BBE)

–An errored block not occurring as a part of an SES

•Unavailable seconds (UAS)

–A period of unavailable time begins at the onset of ten consecutive SES events. These ten seconds are considered to be part of unavailable time. A new period of available time begins at the onset of 10 consecutive non-SES events. These ten seconds are considered to be part of available time. UAS is the number of second of unavailable time.

This section describes:

• "Present G.826 PM data"

• "View Counters"

8.1.2  Present G.826 PM data

G.826 PM data is available in the network elements under management.

8.1.2.1  Background

The network elements have limited memory for historical data storage, and the oldest data will be removed in favour of new, current data registered to the network element¹.

G.826 specifies the accumulated performance data for 15 minutes and 24 hours periods. The incomplete data for the current 15 minutes period is also available. This is updated continuously on the network element. PM data is stored in time series measurement periods (for instance such as the last 15 minutes and the last 24 hours period) on the network element.

From Management Tree, you can select G.826 PM data for the following managed objects:

Table 8-1 Managed Objects

Parameter

Description

Objects

RS

Regenerator Section

near end

MS

Multiplex Section

near end and far end

VC-3

Virtual Container 3

near end and far end

VC-4

Virtual Container 4

near end and far end

VC-12

Virtual Container 12

near end and far end

•None- intrusive monitors on AU's and TUs

Near End and Far End data apply to all managed object, except for the RS.

Figure 8-1 View PM - Example

With near end and far end data for all managed object except the RS. The managed objects have PM attributes as defined in the information model, Figure 8-1.

Available time periods are:

•15 minutes

•24 hours

The system presents current data and historical data. The number of historical stored periods are:

•16x15 minutes

•1x24 hours

See also "5.4.7  Monitoring PDH Port VC-n Performance" section on page 5-12 and "5.15.10  Monitoring WAN Port Performance" section on page 5-66.

8.1.3  View Counters

From the Management Tree you can view all managed objects that are monitored objects, for instance:

• LAN ports

• WAN ports

• Bridge

• IP

• IPM

• IPX

8.1.4  Criteria for Counting Valid-data

Criteria for PM counters: for disabled ports, there is no PM-counting (all BBE, ES, SES, UAS have value 0), valid-data flag not affected (data is set as valid if conditions mentioned below are fulfilled).

For Valid-data flag (used for previous 15-min/24-hour intervals) the following rules apply:

The flag will not be set for any 15-min period (for any levels) if 600 seconds (10 minutes) or less are counted (since counter-reset or device-reset).

•The flag will not be set for any 24-hour period (for any levels) if 20 hours or less are counted (since counter-reset or device-reset).

•For RS/MS/VC-4 levels the flag will not be set for any 15-min if the STM-n port was not defined at the beginning of the period (defined meaning that STM-n port was expected in that slot/port position).

•For RS/MS/VC-4 levels the flag will not be set for any 24-hour period if rule 3) was true for 80 15-min intervals (20 hours) or less.

•For VC-4 level the flag will not be set for any 15-min period if the AUG-1 is not structured as AUG_AU4_TO_XC or AUG_TUG3x3 at the beginning of the period.

•For VC-4 level the flag will not be set for any 24-hour period if rule 5) was true for 80 15-min intervals (20 hours) or less.

In all other cases the valid-data flags are set to

•RS-level valid-data: rules 1,2,3,4

•MS-level valid-data: rules 1,2,3,4

•VC-4-level valid-data: rules 1,2,3,4,5,6

•E3-(/VC-3-) valid-data: rules 1,2

•E1(/VC-12-) valid-data: rules 1,2

•WAN (/VC-12-) valid-data: rules 1,2

8.2  Manage RMON

This section describes management of RMON (Remote Network Monitoring) within a network of Cisco network elements.

RMON is an IETF standard (RFC 2819) for monitoring the status of a network by activating probes/monitors at targeted ports in the network, and using RMON clients to collect and present the status information.

8.2.1  AboUt Rmon Measurements In Cisco Network Elements

With Cisco RMON Management, the user can set up local monitoring devices (probes) at selected locations in the managed network. The RMON monitors perform data acquisition and local storage.

An RMON monitor can be set up to accumulate information (perform diagnostics and collect statistics) in local PM buffer, and to log PM associated events or to generate RMON traps to the management system. To collect information, the management system can either poll the monitors regularly or be notified by monitor notifications (SNMP traps), triggered by exception conditions.

8.2.2  RMON Overview

Manage time series measurements on local area networks (LANs) and interconnecting E1/E3 lines from a central site with RMON. RMON uses monitoring probes to acquire and store measurements. Remote monitoring according to RMON needs management, and has to be set up specifically.

With RMON management you create RMON monitors, configuring measurement sessions, alarm conditions, logging and alarm generation, and acquire RMON data views.

•Events is used to define event types generated from the device.

• Alarms is used to set up alarm thresholds for RMON monitoring.

•History Control is used to set up the periodic sampling of information from the network element ports, and to configure RMON probes/ monitors.

•Logs is used to log selected events from RMON Events.

• Statistics contains statistics on each monitored ethernet interface on the network element.

Figure 8-2 RMON - GUI overview

The following sub- sections explain how to create, configure and inspect RMON monitor(s):

• "Create RMON History Monitor(s)".

• "Create RMON Event Monitor".

• "View RMON Data".

8.2.3  Create RMON Event Monitor

This section describes how to create an RMON Event Monitor as follows:

• "Define RMON Event Types".

• "Configure an RMON Event Monitor"

– "Define A Monitor Source"

– "Define detection criteria"

8.2.3.1  Define RMON Event Types

Set RMON event types prior to configuring RMON Event Monitors. Events are defined as entries in the RMON Event Table. You can choose between the following alternatives:

•Generate log entries (to RMON Logs)

•Generate RMON traps (to Event Trace, Notification List)

•A combination of Log entry and RMON traps
(as illustrated in the following sections)

Step 1 Add a new event to RMON Event Table.

Step 2 Select event Type (here: Log entry and SNMP trap).

Step 3 Enter Community string for communication with intended trap recipient.

Figure 8-3

RMON Events-GUI example

Step 4 Enter an Owner (optional).

Step 5 Enter a Description (optional).

Step 6 Repeat the procedure for as many event definitions as desired.

Step 7 Save event definition.

Note The event Id is used in configuring RMON event monitors.

8.2.4  Configure an RMON Event Monitor

You configure RMON Event Monitor by defining the conditions for event generation from RMON Alarms. If editing an existing monitor, you select the desired table entry.

8.2.4.1  Define A Monitor Source

Step 1 Select Alarms from RMON attributes.

Figure 8-4 Select RMON Alarms

Step 2 Add a new monitor entry to RMON Alarm table.

Figure 8-5 Add Monitor Entry

Step 3 Specify a LAN/ WAN port as DataSource for desired ethernet interface instance (Port if-index or BridgePortNumber.)

Step 4 Enter an Owner for monitor administration (optional).

Step 5 Select the port status to be used in detecting the alarm situation from AlarmVariable pull- down list.

Define detection criteria

Step 6 Set Interval for the periodic check for alarm conditions.

Step 7 Choose absolute values or delta values from Sample Type if the monitor checks alarm conditions (difference between last and current sample).

Step 8 Set the desired alarm Startup conditions:

This attribute defines the initial behavior of the monitor:
Alarms are generated if the monitor is in an alarm situation (below falling threshold or above rising threshold) at startup.

Alarm Rising Threshold (integer - number of occurrences).
The monitor triggers an event when the counter value rises above this level.

Alarm Falling Threshold (integer - number of occurrences).
The monitor triggers an event when the counter value sinks below this level.

Rising Event cross-reference.
A cross reference to the Event Table that identifies the event that is triggered when the monitor registers a rising threshold violation.

Falling Event cross-reference.
A cross reference to the Event Table that identifies the event that is triggered when the monitor registers a falling threshold violation.

Step 9 Save RMON event monitor configuration.

The network elements RMON monitor can start generating alarm events for selected ethernet interface.

8.2.5  Create RMON History Monitor(s)

Sampling interval is as defined by the user. Sampling duration is as granted by the network element. Each sample contains the full set of RMON counter data associated with the data source.

An RMON historical monitor maintains both a 'current' and a 'historical' list of statistical data. History Statistics contains the periodical samples from one specific Ethernet interface instance.

Step 1 Add a monitor for time series measurements in RMON History Control.

Figure 8-6

RMON History Monitor- example

Step 2 Specify a LAN/ WAN port as DataSource for desired ethernet interface instance (Port if-index or BridgePortNumber.)

Step 3 Enter an Owner (optional).

Step 4 Set polling parameters for each data sampling:
Poll interval in seconds, or poll duration as number of polls in BucketRequestedCount.

Step 5 Save RMON monitor configuration.
The network element starts monitoring historical data.

Step 6 Inspect the poll duration of the RMON monitor from BucketGrantedCount.

Note The monitored data is based on free-running sets of counters. RMON has no support for detecting counter overflow.

8.3  View RMON Data

RMON data is available for inspection in several RMON tables within the network element:

•RMON Statistics present statistics for each monitored ethernet interface. See "Inspection Of Current Statistical Data" on page -9.

•RMON History Statistics present time series of statistics measurements. See "Inspection of History Statistics Per Port" on page -10.

•RMON Logs contain logged events from each monitored interface.
See "View logged events" on page -10.

•RMON Traps are presented in CiscoEdgeCraft Event Trace view. This information is not stored on the network element. See "Notification" on page -19.

You can manage several RMON Monitors in the network element at any time.

8.3.1  View Statistical Data

This section describes the following:

• "Inspection Of Current Statistical Data".

• "Inspection of History Statistics Per Port".

Presented RMON data can be printed to file.

8.3.1.1  Inspection Of Current Statistical Data

Step 1 Press RMON in Management Tree.

Step 2 Select Statistics from RMON attributes.

Figure 8-7

RMON Statistics- GUI overview

The current statistical information of all monitored interfaces for all monitored ports will be presented.

Step 3 Customize visible columns from right- click menu (optional).

8.3.1.2  Inspection of History Statistics Per Port

Step 1 Select History Control from RMON attributes. RMON statistical time series table view for all history monitors and to all the ports is presented.

Figure 8-8

Select History Control

Step 2 Select the RMON history monitor for LAN/ WAN interface you want to inspect.

Step 3 Press History Statistics link to view the RMON statistical time series for that specific port.

Figure 8-9

RMON History statistics - example

8.3.2  View logged events

How to view logged events:

8.3.2.1  Inspection of the Event Log

If the event monitor is configured to log events locally, you can inspect this in the RMON Logs.

Step 1 Select Logs from RMON attributes.

All events logged by all monitors within the selected network node will be presented.

Figure 8-10

RMON Logs-view

8.3.2.2  Inspection of the Event Log

Inspect events filtered for a specific LAN/ WAN port.

Step 1 Select desired RMON monitor from RMON Alarms.

Step 2 Open RMON Event logs from Rising Event and/ or Falling events.

Figure 8-11

Alarm monitor- example

A filtered table view of the Event Log will be presented, containing all events associated by the selected monitor. All logged events are presented with timestamp and description fields.

Figure 8-12

RMON Event log to alarm monitor-view

8.3.2.3  Delete RMON Monitor

Step 1 Select desired monitor (Alarms or History) in the relevant RMON table.

Step 2 Press Delete.

RMON monitor (and any RMON event definitions) is removed in from the network element.

RMON data presentation is closed on your desktop. When deleting RMON monitors, the associated log table entries (History Statistics, Event Log entries) are removed automatically.

¹Time periods and interval may vary between the different network element types.

Parameter	Description	Objects
RS	Regenerator Section	near end
MS	Multiplex Section	near end and far end
VC-3	Virtual Container 3	near end and far end
VC-4	Virtual Container 4	near end and far end
VC-12	Virtual Container 12	near end and far end