The ENTITY-MIB supports the Cisco compliance effort for a unique device identifier (UDI) standard stored in Identification Programmable Read-Only Memory (IDPROM).
The Cisco UDI provides a unique identity for every Cisco product. The UDI is composed of three separate data elements that must be stored in the entPhysicalTable:
Orderable product identifier (PID)—The alphanumeric identifier used by customers to order Cisco products. Two examples include A9K-RSP-4G and A9K-4T-E. PID is limited to 18 characters and must be stored in the entPhysicalModelName object.
Version identifier (VID)— The version of the PID. The VID indicates the number of times a product has versioned in ways that are reported to a customer. For example, the product identifier A9K-RSP-4G may have a VID of V04. VID is limited to three alphanumeric characters and must be stored in the entPhysicalHardwareRev object.
Serial number (SN)—The 11-character identifier used to identify a specific part within a product and must be stored in the entPhysicalSerialNum object. Serial number content is defined by manufacturing part number 7018060-0000. The SN is accessed at the following website by searching on the part number 701806-0000:
Note The Version ID returns NULL for those old or existing cards with IDPROMs that do not have the Version ID field. Therefore, corresponding entPhysicalHardwareRev returns NULL for cards that do not have the Version ID field in IDPROM.
Cisco Redundancy Features
Redundancy creates a duplication of data elements and software functions to provide an alternative in case of failure. The goal of Cisco redundancy features is to cut over without affecting the link and protocol states associated with each interface and continue packet forwarding. The state of the interfaces and subinterfaces is maintained, along with the state of line cards and various packet processing hardware.
This section describes the levels of redundancy supported on the Cisco ASR 9000 Series router and how to verify that this feature is available. The Cisco ASR 9000 Series routers supports fault resistance by allowing a Cisco redundant Route Switch Processor (RSP) to take over if the active RSP fails. Redundancy prevents equipment failures from causing service outages, and supports hitless maintenance and upgrade activities. The state of the interfaces and subinterfaces is maintained along with the state of line cards and various packet processing hardware.
Redundant systems support two RSP. One acts as the active RSPs while the other acts as the standby RSPs.
The redundancy feature provides high availability for the Cisco routers by switching when one of the following conditions occur:
Cisco IOS XR Software failure
The Cisco ASR 9000 Series routers operates in Nonstop Forwarding/Stateful Switchover (NSF/SSO) mode.
Nonstop Forwarding/Stateful Switchover
This section describes the Nonstop Forwarding/Stateful Switchover mode. With NSF/SSO, the Cisco ASR 9000 Series routers can change from the active to the standby RSPs almost immediately while continuing to forward packets. Cisco IOS XR Software NSF/SSO support on this platform enables immediate failover.
In networking devices running NSF/SSO, both RSPs must be running the same configuration so that the standby RSP is always ready to assume control following a fault on the active RSP. The configuration information is synchronized from the active RSP to the standby RSP at startup and each timechanges to the active RSP configuration occur.
Following an initial synchronization between the two RSPs, NFS/SSO maintains RSP state information between them, including forwarding information.
The Cisco Nonstop Forwarding (NSF) works with Stateful Switchover (SSO) to minimize the amount of time a network is unavailable to its users following a Route Switching Processor (RSP) fail-over in a router with dual RSPs. NSF/SSO capability allows routers to detect a switchover and take the necessary actions to continue forwarding network traffic and to recover route information from peer devices.
The Cisco NSF works with the Stateful Switchover (SSO) feature in Cisco IOS XR Software to minimize the amount of time a network is unavailable to its users following a switchover. The main objective of the Cisco NSF/SSO is to continue forwarding data packets along known routes while the routing protocol information is restored following a route switchover.
Verifying the Cisco ASR 9000 Series Router Redundancy
To display information about the active and standby RSP engines installed in the Cisco ASR 9000 Series router, use the show redundancy command and show redundancy states command.
Fri Feb 20 01:15:10.213 PST PST
Redundancy information for node 0/RSP0/CPU0:
Node 0/RSP0/CPU0 is in ACTIVE role
Partner node (0/RSP1/CPU0) is in STANDBY role
Standby node in 0/RSP1/CPU0 is ready
Standby node in 0/RSP1/CPU0 is NSR-ready
Reload and boot info
A9K-RSP-4G reloaded Thu Feb 19 09:29:24 2009: 15 hours, 45 minutes ago
Active node booted Thu Feb 19 10:40:02 2009: 14 hours, 35 minutes ago
Last switch-over Thu Feb 19 21:45:59 2009: 3 hours, 29 minutes ago
Standby node boot Thu Feb 19 21:46:57 2009: 3 hours, 28 minutes ago
Standby node last went not ready Thu Feb 19 21:49:06 2009: 3 hours, 26 minutes ago
Standby node last went ready Thu Feb 19 21:49:06 2009: 3 hours, 26 minutes ago
Standby node last went not NSR-ready Thu Feb 19 21:49:27 2009: 3 hours, 25 minutes ago
Standby node last went NSR-ready Thu Feb 19 21:49:27 2009: 3 hours, 25 minutes ago
There have been 2 switch-overs since reload
Managing Physical Entities
This section describes how to use SNMP to manage the physical entities (components) in the router by:
The physical entity management feature of the Cisco ASR 9000 Series router SNMP implementation does the following:
Monitors and configures the status of field-replaceable units (FRUs)
Provides information about physical port to interface mappings
Provides asset information for asset tagging
Provides firmware and software information for chassis components
MIBs Used for Physical Entity Management
CISCO-ENTITY-ASSET-MIB—Contains asset tracking information (IDPROM contents) for the physical entities listed in the entPhysicalTable of the ENTITY-MIB. The MIB provides device-specific information for physical entities, including orderable part number, serial number, manufacturing assembly number, and hardware, software, and firmware information.
CISCO-ENTITY-FRU-CONTROL-MIB—Contains objects used to monitor and configure the administrative and operational status of field-replaceable units (FRUs), such as fans, RSPs, and transceivers that are listed in the entPhysicalTable of the ENTITY-MIB.
CISCO-ENTITY-SENSOR-MIB—Contains information about entities in the entPhysicalTable with an entPhysicalClass value of sensor.
ENTITY-MIB—Contains information for managing physical entities on the router. It also organizes the entities into a containment tree that depicts their hierarchy and relationship to each other. The MIB contains the following tables:
– The entPhysicalTable describes each physical component (entity) in the router. The table contains an entry for the top-level entity (the chassis) and for each entity in the chassis. Each entry provides information about that entity: its name, type, vendor, and a description, and a description of how the entity fits into the hierarchy of chassis entities.
Each entity is identified by a unique index ( entPhysicalIndex ) that is used to access information about the entity in this and other MIBs.
– The entAliasMappingTable maps each physical port’s entPhysicalIndex value to its corresponding ifIndex value in the IF-MIB ifTable.
– The entPhysicalContainsTable shows the relationship between physical entities in the chassis. For each physical entity, the table lists the entPhysicalIndex for each of the entity’s child objects.
Performing Inventory Management
To obtain information about entities in the router, perform a MIB walk on the ENTITY-MIB entPhysicalTable.
As you examine sample entries in the ENTITY-MIB entPhysicalTable, consider the following objects:
entPhysicalIndex—Uniquely identifies each entity in the chassis. This index is also used to access information about the entity in other MIBs.
entPhysicalContainedIn—Indicates the entPhysicalIndex of a component’s parent entity.
entPhysicalParentRelPos—Shows the relative position of same-type entities that have the same entPhysicalContainedIn value (for example, chassis slots and line card ports).
Note The container is applicable if the physical entity class is capable of containing one or more removable physical entities. For example, each (empty or full) slot in a chassis is modeled as a container. All removable physical entities should be modeled within a container entity, such as field-replaceable modules, fans, or power supplies.
Sample of ENTITY-MIB entPhysicalTable Entries
The samples in this section show how information is stored in the entPhysicalTable. You can perform asset inventory by examining entPhysicalTable entries.
Note The sample outputs and values that appear throughout this appendix are examples of data you can view when using MIBs.
The following display shows the ENTITY-MIB entPhysicalTable sample entries:
entPhysicalDescr.186 = 4-Port 10GE Extended Line Card, Requires XFPs
entPhysicalDescr.187 = Ten GigabitEthernet Port
entPhysicalDescr.188 = GigeEthernet XFP container
entPhysicalDescr.190 = Transceiver Temperature Sensor
entPhysicalDescr.191 = Transceiver Tx Power Sensor
entPhysicalDescr.192 = Transceiver Rx Power Sensor
entPhysicalDescr.193 = Transceiver Transmit Bias Current Sensor
entPhysicalDescr.194 = Line Card host
entPhysicalDescr.195 = Inlet Temperature Sensor
entPhysicalDescr.196 = Hot Temperature Sensor
entPhysicalDescr.197 = Voltage Sensor - IBV
entPhysicalDescr.198 = Voltage Sensor - 5.0V
entPhysicalDescr.199 = Voltage Sensor - VP3P3_CAN
entPhysicalDescr.200 = Voltage Sensor - 3.3V
Where entPhysicalDescr identifies the manufacturer name for the physical entity.
Where entPhysicalVendorType identifies the unique vendor-specific hardware type of the physical entity.
entPhysicalContainedIn.186 = 92
entPhysicalContainedIn.187 = 186
entPhysicalContainedIn.188 = 187
entPhysicalContainedIn.189 = 188
entPhysicalContainedIn.190 = 189
entPhysicalContainedIn.191 = 189
entPhysicalContainedIn.192 = 189
entPhysicalContainedIn.193 = 189
entPhysicalContainedIn.194 = 186
entPhysicalContainedIn.195 = 194
entPhysicalContainedIn.196 = 194
entPhysicalContainedIn.197 = 194
entPhysicalContainedIn.198 = 194
entPhysicalContainedIn.199 = 194
entPhysicalContainedIn.200 = 194
Where entPhysicalContainedIn indicates the entPhysicalIndex of a parent entity (component).
entPhysicalClass.186 = module(9)
entPhysicalClass.187 = port(10)
entPhysicalClass.188 = container(5)
entPhysicalClass.189 = module(9)
entPhysicalClass.190 = sensor(8)
entPhysicalClass.191 = sensor(8)
entPhysicalClass.192 = sensor(8)
entPhysicalClass.193 = sensor(8)
entPhysicalClass.194 = module(9)
entPhysicalClass.195 = sensor(8)
entPhysicalClass.196 = sensor(8)
entPhysicalClass.197 = sensor(8)
entPhysicalClass.198 = sensor(8)
entPhysicalClass.199 = sensor(8)
entPhysicalClass.200 = sensor(8)
Where entPhysicalClass indicates the general type of hardware device.
entPhysicalParentRelPos.186 = 0
entPhysicalParentRelPos.187 = 1
entPhysicalParentRelPos.188 = 0
entPhysicalParentRelPos.189 = 0
entPhysicalParentRelPos.190 = 0
entPhysicalParentRelPos.191 = 1
entPhysicalParentRelPos.192 = 2
entPhysicalParentRelPos.193 = 3
entPhysicalParentRelPos.194 = 0
entPhysicalParentRelPos.195 = 0
entPhysicalParentRelPos.196 = 1
entPhysicalParentRelPos.197 = 2
entPhysicalParentRelPos.198 = 3
entPhysicalParentRelPos.199 = 4
entPhysicalParentRelPos.200 = 5
Where entPhysicalParentRelPos indicates the relative position of this child among the other entities.
entPhysicalName.186 = module 0/5/CPU0
entPhysicalName.187 = TenGigE0/5/0/1
entPhysicalName.188 = slot mau 0/5/CPU0/1
entPhysicalName.189 = module mau 0/5/CPU0/1
entPhysicalName.190 = temperature 0/5/CPU0/1
entPhysicalName.191 = power Tx 0/5/CPU0/1
entPhysicalName.192 = power Rx 0/5/CPU0/1
entPhysicalName.193 = current 0/5/CPU0/1
entPhysicalName.194 = module 0/5/CPU0
entPhysicalName.195 = temperature 0/5/CPU0
entPhysicalName.196 = temperature 0/5/CPU0
entPhysicalName.197 = voltage 0/5/CPU0
entPhysicalName.198 = voltage 0/5/CPU0
entPhysicalName.199 = voltage 0/5/CPU0
entPhysicalName.200 = voltage 0/5/CPU0
Where entPhysicalName provides the textual name of the physical entity.
Where entPhysicalHardwareRev provides the vendor-specific hardware revision number (string) for the physical entity.
entPhysicalFirmwareRev.186 = Version 0.63(20081010:215422)
Where entPhysicalFirmwareRev provides the vendor-specific firmware revision number (string) for the physical entity.
entPhysicalSoftwareRev.186 = 220.127.116.11I
entPhysicalSoftwareRev.189 = 18.104.22.168I
entPhysicalSoftwareRev.194 = 22.214.171.124I
Where entPhysicalSoftwareRev provides the software revision number for the physical entity.
entPhysicalSerialNum.186 = FHH1213002A
entPhysicalSerialNum.189 = ECL114704JD
Where entPhysicalSerialNum provides the vendor-specific serial number (string) for the physical entity.
entPhysicalMfgName.186 = Cisco Systems Inc.
Where entPhysicalMfgName provides the manufacturer name for the physical component.
entPhysicalModelName.186 = A9K-4T-E
entPhysicalModelName.189 = ONS-XC-10G-S1
Where entPhysicalModelName provides the vendor-specific model name string for the physical component.
entPhysicalAlias.194 = host
Where entPhysicalAlias provides the alias name for the physical component.
Where entPhysicalAssetID provides the vendor-specific asset ID for the physical component.
entPhysicalIsFRU.186 = true(1)
entPhysicalIsFRU.187 = false(2)
entPhysicalIsFRU.188 = false(2)
entPhysicalIsFRU.189 = true(1)
entPhysicalIsFRU.190 = false(2)
entPhysicalIsFRU.191 = false(2)
entPhysicalIsFRU.192 = false(2)
entPhysicalIsFRU.193 = false(2)
entPhysicalIsFRU.194 = false(2)
entPhysicalIsFRU.195 = false(2)
entPhysicalIsFRU.196 = false(2)
entPhysicalIsFRU.197 = false(2)
entPhysicalIsFRU.198 = false(2)
entPhysicalIsFRU.199 = false(2)
entPhysicalIsFRU.200 = false(2)
Where entPhysicalIsFRU indicates whether or not this physical entity is considered a field-replaceable unit (FRU).
Note the following about the sample configuration:
All chassis slots and line card ports have the same entPhysicalContainedIn value:
– For chassis slots, entPhysicalContainedIn = 1 (the entPhysicalIndex of the chassis).
– For line card ports, entPhysicalContainedIn = 26 (the entPhysicalIndex of the line card).
Each chassis slot and line card port has a different entPhysicalParentRelPos to show its relative position within the parent object.
Determining the ifIndex Value for a Physical Port
The ENTITY-MIB entAliasMappingIdentifier maps a physical port to an interface by mapping the port’s entPhysicalIndex to its corresponding ifIndex value in the IF-MIB ifTable. The following sample shows that the physical port with a entPhysicalIndex value of 35 is associated with the interface with the ifIndex value of four:
entAliasMappingIdentifer.35.0 = ifIndex.4
Note See the MIB for detailed descriptions of possible MIB values.
Monitoring and Configuring FRU Status
View objects in the CISCO-ENTITY-FRU-CONTROL-MIB cefcModuleTable to determine the administrative and operational status of FRUs, such as power supplies and line cards:
cefcModuleAdminStatus—The administrative state of the FRU. This object is read-only and returns enable.
cefcModuleOperStatus—The current operational state of the FRU.
Figure 1-1 shows a cefcModuleTable entry for a line card with the entPhysicalIndex value of 24.
You can use the CLI or SNMP to identify hosts to receive SNMP notifications and to specify the types of notifications they are to receive (notifications). For CLI instructions, see the “Enabling Notifications” section. To use SNMP to configure this information:
Use SNMP-NOTIFICATION-MIB objects, including the following, to select target hosts and specify the types of notifications to generate for those hosts:
snmpNotifyTable—Contains objects to select hosts and notification types:
– snmpNotifyTag is an arbitrary octet string (a tag value) used to identify the hosts to receive SNMP notifications. Information about target hosts is defined in the snmpTargetAddrTable (SNMP-TARGET-MIB), and each host has one or more tag values associated with it. If a host in snmpTargetAddrTable has a tag value that matches this snmpNotifyTag value, the host is selected to receive the types of notifications specified by snmpNotifyType.
– snmpNotifyType is the type of SNMP notification to send: notification(1) or inform(2).
snmpNotifyFilterProfileTable and snmpNotifyFilterTable—Use objects in these tables to create notification filters to limit the types of notifications sent to target hosts.
Use SNMP-TARGET-MIB objects to configure information about the hosts to receive notifications:
snmpTargetAddrTable—Transport addresses of hosts to receive SNMP notifications. Each entry provides information about a host address, including a list of tag values:
– snmpTargetAddrTagList—A set of tag values associated with the host address. If a host tag value matches snmpNotifyTag, the host is selected to receive the types of notifications defined by snmpNotifyType.
snmpTargetParamsTable—SNMP parameters to use when generating SNMP notifications.
Use the notification enable objects in appropriate MIBs to enable and disable specific SNMP notifications.
If entity notifications are enabled, the router generates an entConfigChange notification (ENTITY-MIB) when the information in any of the following tables changes (which indicates a change to the router configuration):
Note A management application that tracks configuration changes checks the value of the entLastChangeTime object to detect any entConfigChange notifications that were missed as a result of throttling or transmission loss.
Enabling Notifications for Configuration Changes
To configure the router to generate an entConfigChange notification each time its configuration changes, enter the snmp-sever trap entity command from the CLI. Use the no form of the command to disable the notifications.
Router(config)# snmp-server traps entity
Router(config)# no snmp-server traps entity
FRU Status Changes
If FRU notifications are enabled, the router generates the following notifications in response to changes in the status of a FRU:
cefcModuleStatusChange—The operational status (cefcModuleOperStatus) of a FRU changes.
cefcFRUInserted—A FRU is inserted in the chassis. The notification indicates the entPhysicalIndex of the FRU and the container in which it was inserted.
cefcFRURemoved—A FRU is removed from the chassis. The notification indicates the entPhysicalIndex of the FRU and the container from which it was removed.
Note See the CISCO-ENTITY-FRU-CONTROL-MIB for more information about these notifications.
Enabling FRU Notifications
To configure the router to generate notifications for FRU events, enter the snmp-server traps fru-ctrl command from the CLI. Use the no form of the command to disable the notifications.
Router(config)# snmp-server traps fru-ctrl
Router(config)# no snmp-server traps fru-ctrl
To enable FRU notifications through SNMP, set cefcMIBEnableStatusNotification to true(1). Disable the notifications by setting cefcMIBEnableStatusNotification to false(2).
Monitoring Quality of Service
This section provides the following information about using Quality of Service (QoS) in your configuration:
The Cisco ASR 9000 Series router distributes QoS features across the line cards. Line cards are designed to provide QoS features on packets that flow through the line cards.
The CISCO-CLASS-BASED-QOS-MIB provides read-only access to Quality of Service (QoS) configuration information and statistics for Cisco platforms that support the modular Quality of Service command-line interface (modular QoS CLI).
CISCO-CLASS-BASED-QOS-MIB Object Relationship
To understand how to navigate the CISCO-CLASS-BASED-QOS-MIB tables, it is important to understand the relationship among different QoS objects. QoS objects consists of:
Match statement—Specific match criteria to identify packets for classification purposes.
Class map—A user-defined traffic class that contains one or more match statements used to classify packets into different categories.
Feature action – Action taken on classified traffic. Features include police, traffic shaping, queueing, random detect, and packet marking. After the traffic is classified actions are applied to packets matching each traffic class.
Policy map – A user-defined policy that associates QoS feature actions to user-defined class maps as policy maps can have multiple class maps.
Service policy—A policy map that has been attached to an interface.
The MIB uses the following indices to identify QoS features and distinguish among instances of those features:
cbQosObjectsIndex – Identifies each QoS feature on the router.
cbQoSConfigIndex – Identifies a type of QoS configuration. This index is shared by QoS objects that have identical configurations.
cbQosPolicyIndex – Identifies a unique service policy.
QoS MIB Information Storage
CISCO-CLASS-BASED-QOS-MIB information is stored as:
Configuration information— Includes all the QoS configuration objects, such as class maps, policy map, match statements, and feature action configuration parameters. The configuration may have multiple identical instances. Configuration objects are identified by cbQosConfigIndex attribute. Multiple instances of the same QoS feature share a single configuration object that is identified by the same cbQosConfigIndex value.
Service-policy information— Includes instances of all QoS objects, such as service-policies, classes, match statements, and feature actions. Service-policies are identified by cbQosPolicyIndex and instances of QoS objects are identified by the combination of cbQosPolicyIndex and cbQosObjectsIndex attributes.
QoS Hardware Configuration and Statistic Support
The CISCO-CLASS-BASED-QOS-MIB does not cover all the Cisco ASR 9000 Series router QoS hardware configuration and statistics.
The Cisco ASR 9000 Series router supports the concept of ‘shared policy instance’ where, based on the configuration, the resources for individual service policies are shared among multiple interfaces. The cbQosMIB attribute does not indicate whether the service-policies are shared-policy instances or non-shared policy instances.
The interfaces associated with the shared policy instance have a separate entry in the cbQosServicePolicyTable. The MIB entries, associated with each interface that is a part of the same shared-policy-instance, have the same data values, for example, everything except for the cbQosServicePolicyTable is identical for the rows associated with the values of cbQosPolicyIndex for such interfaces.
Figure 1-2 shows how the indexes provide access to QoS configuration information and statistics.
Figure 1-2 The Cisco ASR 9000 Series Router QoS Indexes
Accessing QoS Configuration Information
To access QoS configuration information and statistics for a particular QoS feature:
Step 1 Look in cbQosServicePolicyTable and find the cbQosPolicyIndex assigned to the policy in which the feature is used.
Step 2 Use cbQosPolicyIndex to access the cbQosObjectsTable, and find the cbQosObjectsIndex and cbQosConfigIndex assigned to the QoS feature.
a. Use cbQosConfigIndex to access configuration tables (cbQosxxxCfgTable) for information about the QoS feature.
b. Use cbQosPolicyIndex and cbQosObjectsIndex to access QoS statistics tables (cbQosxxxStatsTable) for information about the QoS feature.
Viewing QoS Configuration Settings Using the CISCO-CLASS-BASED-QOS-MIB
This section contains an example that shows how QoS configuration settings are stored in CISCO-CLASS-BASED-QOS-MIB tables. The sample shows information grouped by QoS object; however, the actual output of an SNMP query might show QoS information similar to the following.
Note This is only a partial display of all QoS information.
ASR 9000# getmany -v3 10.86.0.94 test-user ciscoCBQosMIB CbQosServicePolicyTable
cbQosIfType.1047 = subInterface(2)
cbQosIfType.1052 = subInterface(2)
cbQosPolicyDirection.1047 = input(1)
cbQosPolicyDirection.1052 = output(2)
cbQosIfIndex.1047 = 36
cbQosIfIndex.1052 = 36
cbQosFrDLCI.1047 = 0
cbQosFrDLCI.1052 = 0
cbQosAtmVPI.1047 = 0
cbQosAtmVPI.1052 = 0
cbQosAtmVCI.1047 = 0
cbQosAtmVCI.1052 = 0
cbQosConfigIndex.1047.1047 = 1045
cbQosConfigIndex.1047.1048 = 1025
cbQosConfigIndex.1047.1050 = 1027
cbQosConfigIndex.1047.1051 = 1046
cbQosConfigIndex.1052.1052 = 1045
cbQosConfigIndex.1052.1053 = 1025
cbQosConfigIndex.1052.1055 = 1027
cbQosConfigIndex.1052.1056 = 1046
cbQosObjectsType.1047.1047 = policymap(1)
cbQosObjectsType.1047.1048 = classmap(2)
cbQosObjectsType.1047.1050 = matchStatement(3)
cbQosObjectsType.1047.1051 = police(7)
cbQosObjectsType.1052.1052 = policymap(1)
cbQosObjectsType.1052.1053 = classmap(2)
cbQosObjectsType.1052.1055 = matchStatement(3)
cbQosObjectsType.1052.1056 = police(7)
cbQosParentObjectsIndex.1047.1047 = 0
cbQosParentObjectsIndex.1047.1048 = 1047
cbQosParentObjectsIndex.1047.1050 = 1048
cbQosParentObjectsIndex.1047.1051 = 1048
cbQosParentObjectsIndex.1052.1052 = 0
cbQosParentObjectsIndex.1052.1053 = 1052
cbQosParentObjectsIndex.1052.1055 = 1053
cbQosParentObjectsIndex.1052.1056 = 1053
cbQosPolicyMapName.1045 = pm-1Meg
cbQosCMName.1025 = class-default
cbQosCMInfo.1025 = matchAny(3)
. . .
Monitoring QoS Using the CISCO-CLASS-BASED-QOS-MIB
This section describes how to monitor QoS on the router by checking the QoS statistics in the CISCO-CLASS-BASED-QOS-MIB tables.
Note The CISCO-CLASS-BASED-QOS-MIB may contain more information than what is displayed in the output of CLI show commands.
Table 1-1 lists the types of QoS statistics tables.
Table 1-1 QoS Statistics Tables
Class map—Counts of packets, bytes, and bit rate before and after QoS policies are executed. Counts of dropped packets and bytes.
Police action—Counts of packets, bytes, and bit rate that conforms to, exceeds, and violates police actions.
Queueing—Counts of discarded packets and bytes, and queue depths.
Traffic shaping—Counts of delayed and dropped packets and bytes, the state of a feature, and queue size.
Random early detection—Counts of packets and bytes dropped when queues are full, and counts of bytes and octets transmitted.
Considerations for Processing QoS Statistics
The router maintains 64-bit counters for most QoS statistics. However, some QoS counters are implemented as a 32-bit counter with a 1-bit overflow flag. In the following samples, the counters are shown as 33-bit counters.
When accessing QoS counter statistics, consider the following:
SNMPv2c or SNMPv3 applications—Access the entire 64 bits of the QoS counter through cbQosxxx64 MIB objects.
SNMPv1 applications—Access QoS statistics in the MIB as follows:
– Access the lower 32 bits of the counter through cb Qosxxx MIB objects.
– Access the upper 32 bits of the counter through cbQosxxxOverflow MIB objects.
Sample QoS Statistics Tables
The samples in this section show the counters in CISCO-CLASS-BASED-QOS-MIB statistics tables:
Figure 1-3 shows the counters in the cbQosCMStatsTable and the indexes for accessing these and other statistics.
Figure 1-4 shows the counters in cbQosMatchStmtStatsTable, cbQosPoliceStatsTable, cbQosQueueingStatsTable, cbQosTSStatsTable, and cbQosREDClassStatsTable.
For ease-of-use, the following figures show some counters as a single object even though the counter is implemented as three objects. For example, cbQosCMPrePolicyByte is implemented as:
Figure 1-3 QoS Class Map Statistics and Indexes
Figure 1-4 QoS Statistics Tables
Sample QoS Applications
This section presents examples of code showing how to retrieve information from the CISCO-CLASS-BASED-QOS-MIB to use for QoS billing operations. You can use the examples to help you develop billing applications. The topics include:
This section describes a sample algorithm that checks the CISCO-CLASS-BASED-QOS-MIB for customer interfaces with service policies, and marks those interfaces for further application processing (such as billing for QoS services).
The algorithm uses two SNMP get-next requests for each customer interface. For example, if the router has 2000 customer interfaces, 4000 SNMP get-next requests are required to determine if those interfaces have transmit and receive service policies associated with them.
Note This algorithm is for informational purposes only. Your application needs may be different.
Check the MIB to see which interfaces are associated with a customer. Create a pair of flags to show if a service policy has been associated with the transmit and receive directions of a customer interface. Mark noncustomer interfaces TRUE (so no more processing is required for them).
Examine the cbQosServicePolicyTable and mark each customer interface that has a service policy attached to it. Also note the direction of the interface.
x = 0;
done = FALSE;
status = snmp-getnext (
ifIndex = cbQosIfIndex.x,
direction = cbQosPolicyDirection.x
IF (status != ‘noError’) THEN
done = TRUE
x = extract cbQosPolicyIndex from response;
IF (direction == ‘output’) THEN
servicePolicyAssociated[ifIndex].transmit = TRUE;
servicePolicyAssociated[ifIndex].receive = TRUE;
Manage cases in which a customer interface does not have a service policy attached to it.
FOR each ifEntry DO
IF (!servicePolicyAssociated[ifIndex].transmit) THEN
Perform processing for customer interface without a transmit service policy.
IF (!servicePolicyAssociated[ifIndex].receive) THEN
Perform processing for customer interface without a receive service policy.
Retrieving QoS Billing Information
This section describes a sample algorithm that uses the CISCO-CLASS-BASED-QOS-MIB for QoS billing operations. The algorithm periodically retrieves post-policy input and output statistics, combines them, and sends the result to a billing database.
The algorithm uses the following:
One SNMP get request per customer interface—To retrieve the ifAlias.
Two SNMP get-next requests per customer interface—To retrieve service policy indexes.
Two SNMP get-next requests per customer interface for each object in the policy—To retrieve post-policy bytes. For example, if there are 100 interfaces and 10 objects in the policy, the algorithm requires 2000 get-next requests (2 x 100 x 10).
Note This algorithm is for informational purposes only. Your application needs may be different.
status = snmp-get (bytes = cbQosCMPostPolicyByte64.x.y);
IF (status == ‘noError’)
total += bytes;
Monitoring Router Interfaces
This section provides information about how to monitor the status of router interfaces to see if there is a problem or a condition that might affect service on the interface. To determine if an interface is Down or experiencing problems, you can:
Step 2 View the setting of the ifLinkUpDownTrapEnable object (IF-MIB ifXTable) for each interface to determine if linkUp and linkDown notifications are enabled or disabled for that interface.
Step 3 To enable linkUp and linkDown notifications on an interface, set ifLinkUpDownTrapEnable to enabled(1).
Step 4 To enable the Internet Engineering Task Force (IETF) standard for linkUp and linkDown notifications, issue the snmp-server trap link ietf command. (The IETF standard is based on RFC 2233.)
Router(config)# snmp-server trap link ietf
Step 5 To disable notifications, use the no form of the snmp-server command.
Billing Customers for Traffic
This section describes how to use SNMP interface counters and QoS data information to determine the amount to bill customers for traffic. It also includes a scenario for demonstrating that a QoS service policy attached to an interface is policing traffic on that interface.
Consider the following important information about IF-MIB counter support:
Unless noted, all IF-MIB counters are supported on the Cisco ASR 9000 Series router interfaces.
For IF-MIB high capacity counter support, Cisco conforms to the RFC 2863 standard. The RFC 2863 standard states that for interfaces that operate:
– At 20 million bits per second or less, 32-bit and packet counters must be supported.
– Faster than 20 million bits per second and slower than 650 million bits per second, 32-bit packet counters and 64-bit octet counters must be supported.
– At 650 million bits per second or faster, 64-bit packet counters and 64-bit octet counters must be supported.
When a QoS service policy is attached to an interface, the router applies the rules of the policy to traffic on the interface and increments the packet and byte counts on the interface.
The following CISCO-CLASS-BASED-QOS-MIB objects provide interface counts:
cbQosCMDropPkt and cbQosCMDropByte (cbQosCMStatsTable)—Total number of packets and bytes that were dropped as they exceeded the limits set by the service policy. These counts include only those packets and bytes that were dropped as they exceeded service policy limits. The counts do not include packets and bytes dropped for other reasons.
cbQosPoliceConformedPkt and cbQosPoliceConformedByte (cbQosPoliceStatsTable)—Total number of packets and bytes that conformed to the limits of the service policy and were transmitted.
Determining the Amount of Traffic to Bill to a Customer
Perform the following steps to determine how much traffic on an interface is billable to a particular customer:
Step 1 Determine which service policy on the interface applies to the customer.
Step 2 Determine the index values of the service policy and class map used to define the customer’s traffic. You need this information in the following steps.
Step 3 Access the cbQosPoliceConformedPkt object (cbQosPoliceStatsTable) for the customer to determine the amount of traffic on the interface that is billable to this customer.
Step 4 (Optional) Access the cbQosCMDropPkt object (cbQosCMStatsTable) for the customer to determine how much of the customer’s traffic was dropped as it exceeded service policy limits.
Scenario for Demonstrating QoS Traffic Policing
This section describes a scenario that demonstrates the use of SNMP QoS statistics to determine how much traffic on an interface is billable to a particular customer. It also shows how packet counts are affected when a service policy is applied to traffic on the interface.
To create the scenario, perform the following steps (each step described in the section below):
1. Create and attach a service policy to an interface.
2. View packet counts before the service policy is applied to traffic on the interface.
3. Issue a ping command to generate traffic on the interface. Note that the service policy is applied to the traffic.
4. View packet counts after the service policy is applied to determine how much traffic to bill the customer for:
Conformed packets—The number of packets within the range set by the service policy and for which you can charge the customer.
Exceeded or dropped packets—The number of packets that were not transmitted because they were outside the range of the service policy. These packets are not billable to the customer.
Note In this scenario, the Cisco ASR 9000 Series router is used as an interim device (that is, traffic originates elsewhere and is destined for another device).
Service Policy Configuration
The following example uses policy map configuration.
police 8000 1000 2000 conform-action transmit exceed-action drop
description VLAN voor klant
encapsulation dot1Q 10
ip address 10.0.0.17 255.255.255.248
service-policy output police-out
Packet Counts Before the Service Policy Is Applied
The following CLI and SNMP output shows the output traffic for interface before the service policy is applied:
ASR 9000# getmany -v2c 10.86.0.63 public ciscoCBQosMIB
. . .
cbQosCMDropPkt.1143.1145 = 57
. . .
cbQosPoliceConformedPkt.1143.1151 = 42
. . .
Using IF-MIB Counters
This section describes the IF-MIB counters and how you can use them on various interfaces and subinterfaces. The subinterface counters are specific to the protocols. This section addresses the IF-MIB counters for ATM interfaces.
The IF-MIB counters are defined with respect to lower and upper layers:
ifInDiscards—The number of inbound packets that were discarded, even though no errors were detected to prevent their being deliverable to a higher-layer protocol. One reason for discarding such a packet is to free up buffer space.
IfInErrors—The number of inbound packets that contained errors preventing them from being deliverable to a higher-layer protocol for packet-oriented interfaces.
ifInUnknownProtos—The number of packets received through the interface that were discarded because of an unknown or unsupported protocol for packet-oriented interfaces.
ifOutDiscards—The number of outbound packets that were discarded even though no errors were detected to prevent their being transmitted. One reason for discarding such a packet is to free up buffer space.
ififOutErrors—The number of outbound packets that could not be transmitted because of errors for packet-oriented interfaces.
The logical flow for counters works as follows:
1. When a packet arrives on an interface, check for the following:
a. Error in packet—If any errors are detected, increment ifInErrors and drop the packet.
b. Protocol errors—If any errors are detected, increment ifInUnknownProtos and drop the packet.
c. Resources (buffers)—If unable to get resources, increment ifInDiscards and drop the packet.
d. Increment ifInUcastPkts/ifInNUcastPkts and process the packet (at this point, increment ifInOctets with the size of packet).
2. When a packet is to be sent out of an interface:
a. Increment ifOutUcastPkts/ifOutNUcastPkts (increment ifOutOctets with the size of packet).
b. Check for errors in packet and if there are any errors in packet, increment ifOutErrors and drop the packet.
c. Check for resources (buffers) and if you cannot get resources, then increment ifOutDiscards and drop the packet.
This following output is an example of IF-MIB entries:
The high capacity counters are 64-bit versions of the basic ifTable counters. They have the same basic semantics as their 32-bit counterparts; their syntax is extended to 64 bits.