|
|
Table Of Contents
16.3 SNMP External Interface Requirement
16.6 SNMP Management Information Bases
16.6.1 IETF-Standard MIBs for the ONS 15454
16.6.2 Proprietary ONS 15454 MIBs
16.6.3 Generic Threshold and Performance Monitoring MIBs
16.10.1 64-Bit RMON Monitoring over DCC
16.10.3 Ethernet Statistics RMON Group
16.10.4 History Control RMON Group
16.10.5 Ethernet History RMON Group
SNMP
This chapter explains Simple Network Management Protocol (SNMP) as implemented by the Cisco ONS 15454.
For SNMP setup information, refer to the Cisco ONS 15454 Procedure Guide.
Chapter topics include:
•
SNMP External Interface Requirement
•
16.1 SNMP Overview
SNMP is an application-layer communication protocol that allows ONS 15454 network devices to exchange management information among these systems and with other devices outside the network. Through SNMP, network administrators can manage network performance, find and solve network problems, and plan network growth. Up to 10 SNMP trap destinations and five concurrent CTC user sessions are allowed per node.
The ONS 15454 uses SNMP for asynchronous event notification to a network management system (NMS). ONS SNMP implementation uses standard Internet Engineering Task Force (IETF) management information bases (MIBs) to convey node-level inventory, fault, and performance management information for generic read-only management of DS-1, DS-3, SONET, and Ethernet technologies. SNMP allows a generic SNMP manager such as HP OpenView Network Node Manager (NNM) or Open Systems Interconnection (OSI) NetExpert to be utilized for limited management functions.
The Cisco ONS 15454 supports SNMP Version 1 (SNMPv1) and SNMP Version 2c (SNMPv2c). These versions share many features, but SNMPv2c includes additional protocol operations and 64-bit performance monitoring support. This chapter describes both versions and gives SNMP configuration parameters for the ONS 15454.
Note
Figure 16-1 illustrates the basic layout idea of an SNMP-managed network.
Figure 16-1 Basic Network Managed by SNMP
16.2 Basic SNMP Components
In general terms, an SNMP-managed network consists of a management system, agents, and managed devices.
A management system such as HP OpenView executes monitoring applications and controls managed devices. Management systems execute most of the management processes and provide the bulk of memory resources used for network management. A network might be managed by one or more management systems. Figure 16-2 illustrates the relationship between the network manager, the SNMP agent, and the managed devices.
Figure 16-2 Example of the Primary SNMP Components
An agent (such as SNMP) residing on each managed device translates local management information data, such as performance information or event and error information caught in software traps, into a readable form for the management system. Figure 16-3 illustrates SNMP agent get-requests that transport data to the network management software.
Figure 16-3 Agent Gathering Data from a MIB and Sending Traps to the Manager
The SNMP agent captures data from MIBs, which are device parameter and network data repositories, or from error or change traps.
A managed element—such as a router, access server, switch, bridge, hub, computer host, or network element (such as an ONS 15454)—is accessed through the SNMP agent. Managed devices collect and store management information, making it available through SNMP to other management systems having the same protocol compatibility.
16.3 SNMP External Interface Requirement
Since all SNMP requests come from a third-party application, the only external interface requirement is that a third-part SNMP client application can upload RFC 3273 SNMP MIB variables in the etherStatsHighCapacityTable, etherHistoryHighCapacityTable, or mediaIndependentTable.
16.4 SNMP Version Support
The ONS 15454 supports SNMPv1 and SNMPv2c traps and get requests. The ONS 15454 SNMP MIBs define alarms, traps, and status. Through SNMP, NMS applications can query a management agent for data from functional entities such as Ethernet switches and SONET multiplexers using a supported MIB.
Note
16.5 SNMP Message Types
The ONS 15454 SNMP agent communicates with an SNMP management application using SNMP messages. Table 16-1 describes these messages.
16.6 SNMP Management Information Bases
Section 16.6.1 lists IETF-standard MIBs that are implemented in the ONS 15454 and shows their compilation order. Section 16.6.2 lists proprietary MIBs for the ONS 15454 and shows their compilation order. Section 16.6.3 contains information about the generic threshold and performance monitoring MIBs that can be used to monitor any network element (NE) contained in the network.
16.6.1 IETF-Standard MIBs for the ONS 15454
Table 16-2 lists the IETF-standard MIBs implemented in the ONS 15454 SNMP agents.
First compile the MIBs in Table 16-2. Compile the Table 16-3 MIBs next.
Caution
Table 16-2 IETF Standard MIBs Implemented in the ONS 15454 System
—
IANAifType-MIB.mib
Internet Assigned Numbers Authority (IANA) ifType
1213
RFC1213-MIB-rfc1213.mib
Management Information Base for Network
1907
SNMPV2-MIB-rfc1907.mib
Management of TCP/IP-based Internets: MIB-II
Management Information Base for Version 2 of the Simple Network Management Protocol (SNMPv2)1253
RFC1253-MIB-rfc1253.mib
OSPF Version 2 Management Information Base
1493
BRIDGE-MIB-rfc1493.mib
Definitions of Managed Objects for Bridges
(This defines MIB objects for managing MAC bridges based on the IEEE 802.1D-1990 standard between Local Area Network [LAN] segments.)2819
RMON-MIB-rfc2819.mib
Remote Network Monitoring Management Information Base
2737
ENTITY-MIB-rfc2737.mib
Entity MIB (Version 2)
2233
IF-MIB-rfc2233.mib
Interfaces Group MIB using SNMPv2
2358
EtherLike-MIB-rfc2358.mib
Definitions of Managed Objects for the Ethernet-like Interface Types
2493
PerfHist-TC-MIB-rfc2493.mib
Textual Conventions for MIB Modules Using Performance History Based on 15 Minute Intervals
2495
DS1-MIB-rfc2495.mib
Definitions of Managed Objects for the DS1, E1, DS2 and E2 Interface Types
2496
DS3-MIB-rfc2496.mib
Definitions of Managed Object for the DS3/E3 Interface Type
2558
SONET-MIB-rfc2558.mib
Definitions of Managed Objects for the SONET/SDH Interface Type
2674
P-BRIDGE-MIB-rfc2674.mib Q-BRIDGE-MIB-rfc2674.mib
Definitions of Managed Objects for Bridges with Traffic Classes, Multicast Filtering and Virtual LAN Extensions
3273
HC-RMON-MIB
The MIB module for managing remote monitoring device implementations, augmenting the original RMON MIB as specified in RFC 2819 and RFC 1513 and RMON-2 MIB as specified in RFC 2021
1 RFC = Request for Comment
16.6.2 Proprietary ONS 15454 MIBs
Each ONS 15454 is shipped with a software CD containing applicable proprietary MIBs. Table 16-3 lists the proprietary MIBs for the ONS 15454.
Note
Note
16.6.3 Generic Threshold and Performance Monitoring MIBs
A MIB called CERENT-GENERIC-PM-MIB allows network management stations (NMS) to use a single, generic MIB for accessing threshold and performance monitoring data of different interface types. The MIB is generic in the sense that it is not tied to any particular kind of interface. The MIB objects can be used to obtain threshold values, current performance monitoring (PM) counts, and historic PM statistics for each kind of monitor and any supported interval at the near end and far end.
Previously existing MIBs in the ONS 15454 system provide some of these counts. For example, SONET interface 15-minute current PM counts and historic PM statistics are available using the SONET-MIB. DS-1 and DS-3 counts and statistics are available through the DS1-MIB and DS-3 MIB respectively. The generic MIB provides these types of information and also fetches threshold values and single-day statistics. In addition, the MIB supports optics and dense wavelength division multiplexing (DWDM) threshold and performance monitoring information.
The CERENT-GENERIC-PM-MIB is organized into three different tables:
•
•
•
The cerentGenericPmThresholdTable is used to obtain the threshold values for the monitor types. It is indexed based on the interface index (cerentGenericPmThresholdIndex), monitor type (cerentGenericPmThresholdMonType), location (cerentGenericPmThresholdLocation), and time period (cerentGenericPmThresholdPeriod). The syntax of cerentGenericPmThresholdMonType is type cerentMonitorType, defined in CERENT-TC.mib. The syntax of cerentGenericPmThresholdLocation is type cerentLocation, defined in CERENT-TC.mib. The syntax of cerentGenericPmThresholdPeriod is type cerentPeriod, defined in CERENT-TC.mib.
Threshold values can be provided in 64-bit and 32-bit formats. (For more information about 64-bit counters, see the "HC-RMON-MIB Support" section.) The 64-bit values in cerentGenericPmThresholdHCValue can be used with agents that support SNMPv2. The two 32-bit values (cerentGenericPmThresholdValue and cerentGenericPmThresholdOverFlowValue) can be used by NMSs that only support SNMPv1. The objects compiled in the cerentGenericPmThresholdTable are shown in Table 16-4.
The second table within the MIB, cerentGenericPmStatsCurrentTable, compiles the current performance monitoring (PM) values for the monitor types. The table is indexed based on interface index (cerentGenericPmStatsCurrentIndex), monitor type (cerentGenericPmStatsCurrentMonType), location (cerentGenericPmStatsCurrentLocation) and time period (cerentGenericPmStatsCurrentPeriod). The syntax of cerentGenericPmStatsCurrentIndex is type cerentLocation, defined in CERENT-TC.mib. The syntax of cerentGenericPmStatsCurrentMonType is type cerentMonitor, defined in CERENT-TC.mib. The syntax of cerentGenericPmStatsCurrentPeriod is type cerentPeriod, defined in CERENT-TC.mib.
The cerentGenericPmStatsCurrentTable validates the current PM value using the cerentGenericPmStatsCurrentValid object and registers the number of valid intervals with historical PM statistics in the cerentGenericPmStatsCurrentValidIntervals object.
PM values are provided in 64-bit and 32-bit formats. The 64-bit values in cerentGenericPmStatsCurrentHCValue can be used with agents that support SNMPv2. The two 32-bit values (cerentGenericPmStatsCurrentValue and cerentGenericPmStatsCurrentOverFlowValue) can be used by NMS that only support SNMPv1. The cerentGenericPmStatsCurrentTable is shown in Table 16-5.
The third table in the MIB, cerentGenericPmStatsIntervalTable, obtains historic PM values for the monitor types. This table is indexed based on the interface index, monitor type, location, time period, and interval number. It validates the current PM value in the cerentGenericPmStatsIntervalValid object.
This table is indexed based on interface index (cerentGenericPmStatsIntervalIndex), monitor type (cerentGenericPMStatsIntervalMonType), location (cerentGenericPmStatsIntervalLocation), and period (cerentGenericPmStatsIntervalPeriod). The syntax of cerentGenericPmStatsIntervalIndex is type cerentLocation, defined in CERENT-TC.mib. The syntax of cerentGenericPmStatsIntervalMonType is type cerentMonitor, defined in CERENT-TC.mib. The syntax of cerentGernicPmStatsIntervalPeriod is type cerentPeriod, defined in CERENT-TC.mib.
The table provides historic PM values in 64-bit and 32-bit formats. The 64-bit values contained in the cerentGenericPmStatsIntervalHCValue table can be used with SNMPv2 agents. The two 32-bit values (cerentGenericPmStatsIntervalValue and cerentGenericPmStatsIntervalOverFlowValue) can be used by SNMPv1 NMS. The cerentGenericPmStatsIntervalTable is shown in Table 16-6.
16.7 SNMP Trap Content
The ONS 15454 generates all alarms and events, such as raises and clears, as SNMP traps. These contain the following information:
•
•
•
16.7.1 Generic and IETF Traps
The ONS 15454 supports the generic IETF traps listed in Table 16-7.
16.7.2 Variable Trap Bindings
Each SNMP trap contains variable bindings that are used to create the MIB tables. ONS 15454 traps and variable bindings are listed in Table 16-8. For each group (such as Group A), all traps within the group are associated with all of its variable bindings.
16.8 SNMP Community Names
Community names are used to group SNMP trap destinations. All ONS 15454 trap destinations can be provisioned as part of SNMP communities in Cisco Transport Controller (CTC). When community names are assigned to traps, the ONS 15454 treats the request as valid if the community name matches one that is provisioned in CTC. In this case, all agent-managed MIB variables are accessible to that request. If the community name does not match the provisioned list, SNMP drops the request.
16.9 Proxy Over Firewalls
SNMP and NMS applications have traditionally been unable to cross firewalls used for isolating security risks inside or from outside networks. CTC enables network operations centers (NOCs) to access performance monitoring data such as RMON statistics or autonomous messages across firewalls by using an SNMP proxy element installed on a firewall.
The application-level proxy transports SNMP protocol data units (PDU) between the NMS and NEs, allowing requests and responses between the NMS and NEs and forwarding NE autonomous messages to the NMS. The proxy agent requires little provisioning at the NOC and no additional provisioning at the NEs.
The firewall proxy is intended for use in a gateway network element-end network element (GNE-ENE) topology with many NEs through a single NE gateway. Up to 64 SNMP requests (such as get, getnext, or getbulk) are supported at any time behind single or multiple firewalls. The proxy interoperates with common NMS such as HP OpenView.
For security reasons, the SNMP proxy feature must be enabled at all receiving and transmitting NEs to function. For instructions to do this, refer to the Cisco ONS 15454 Procedure Guide.
16.10 Remote Monitoring
The ONS 15454 incorporates RMON to allow network operators to monitor Ethernet card performance and events. The RMON thresholds are user-provisionable in CTC. Refer to the Cisco ONS 15454 Procedure Guide for instructions. Note that otherwise, RMON operation is invisible to the typical CTC user.
ONS 15454 system RMON is based on the IETF-standard MIB RFC 2819 and includes the following five groups from the standard MIB: Ethernet Statistics, History Control, Ethernet History, Alarm, and Event.
Certain statistics measured on the ML card are mapped to standard MIB if one exists else mapped to a non standard MIB variable. The naming convention used by the standarad/non-standard MIB is not the same as the statistics variable used by the card. Hence when these statistics are obtained via get-reques/get-next-request/SNMP Trap they don't match the name used on the card or as seen by CTC/TL1.
•
16.10.1 64-Bit RMON Monitoring over DCC
The ONS 15454 DCC is implemented over the IP protocol, which is not compatible with Ethernet. The system builds Ethernet equipment History and Statistics tables using HDLC statistics that are gathered over the DCC (running point-topoint protocol, or PPP). RMON DCC (for both IP and Ethernet) monitors the health of remote DCC connections.
RMON DCC contains two MIBS for DCC interfaces. They are:
•
•
16.10.1.1 Row Creation in MediaIndependentTable
The SetRequest PDU for creating a row in the mediaIndependentTable should contain all the values required to activate a row in a single set operation along with an assignment of the status variable to createRequest (2). The SetRequest PDU for entry creation must have all the object IDs (OIDs) carrying an instance value of 0. That is, all the OIDs should be of the type OID.0.
In order to create a row, the SetRequest PDU should contain the following:
•
•
•
The mediaIndependentTable creates a row if the SetRequest PDU is valid according to the above rules. When the row is created, the SNMP agent decides the value of mediaIndependentIndex. This value is not sequentially allotted or contiguously numbered. It changes when an Ethernet interface is added or deleted. The newly created row will have mediaIndependentTable value of valid (1).
If the row already exists, or if the SetRequest PDU values are insufficient or do not make sense, the SNMP agent returns an error code.
Note
The mediaIndependentTable deletes a row if the SetRequest PDU contains a mediaIndependentStatus with a value of invalid (4). The varbind's OID instance value identifies the row for deletion. You can recreate a deleted row in the table if desired.
16.10.1.2 Row Creation in cMediaIndependentHistoryControlTable
SNMP row creation and deletion for the cMediaIndependentHistoryControlTable follows the same processes as for the MediaIndependentTable; only the variables differ.
In order to create a row, the SetRequest PDU should contain the following:
•
•
•
16.10.2 HC-RMON-MIB Support
For the ONS 15454, the implementation of the high-capacity remote monitoring information base (HC-RMON-MIB, or RFC 3273) enables 64-bit support of existing RMON tables. This support is provided with the etherStatsHighCapacityTable and the etherHistoryHighCapacityTable. An additional table, the mediaIndependentTable, and an additional object, hcRMONCapabilities, are also added for this support. All of these elements are accessible by any third-party SNMP client having RFC 3273 support.
16.10.3 Ethernet Statistics RMON Group
The Ethernet Statistics group contains the basic statistics monitored for each subnetwork in a single table called the etherStatsTable.
16.10.3.1 Row Creation in etherStatsTable
The SetRequest PDU for creating a row in this table should contain all the values needed to activate a row in a single set operation, and an assigned status variable to createRequest. The SetRequest PDU object ID (OID) entries must all carry an instance value, or type OID, of 0.
In order to create a row, the SetRequest PDU should contain the following:
•
•
•
The etherStatsTable creates a row if the SetRequest PDU is valid according to the above rules. When the row is created, the SNMP agent decides the value of etherStatsIndex. This value is not sequentially allotted or contiguously numbered. It changes when an Ethernet interface is added or deleted. The newly created row will have etherStatsStatus value of valid (1).
If the etherStatsTable row already exists, or if the SetRequest PDU values are insufficient or do not make sense, the SNMP agent returns an error code.
Note
16.10.3.2 Get Requests and GetNext Requests
Get requests and getNext requests for the etherStatsMulticastPkts and etherStatsBroadcastPkts columns return a value of zero because the variables are not supported by ONS 15454 Ethernet cards.
16.10.3.3 Row Deletion in etherStatsTable
To delete a row in the etherStatsTable, the SetRequest PDU should contain an etherStatsStatus "invalid" value (4). The OID marks the row for deletion. If required, a deleted row can be recreated.
16.10.3.4 64-Bit etherStatsHighCapacity Table
The Ethernet statistics group contains 64-bit statistics in the etherStatsHighCapacityTable, which provides 64-bit RMON support for the HC-RMON-MIB. The etherStatsHighCapacityTable is an extension of the etherStatsTable that adds 16 new columns for performance monitoring data in 64-bit format. There is a one-to-one relationship between the etherStatsTable and etherStatsHighCapacityTable when rows are created or deleted in either table.
16.10.4 History Control RMON Group
The History Control group defines sampling functions for one or more monitor interfaces in the historyControlTable. The values in this table, as specified in RFC 2819, are derived from the historyControlTable and etherHistoryTable.
16.10.4.1 History Control Table
The RMON is sampled at one of four possible intervals. Each interval, or period, contains specific history values (also called buckets). Table 16-9 lists the four sampling periods and corresponding buckets.
The historyControlTable maximum row size is determined by multiplying the number of ports on a card by the number of sampling periods. For example, an ONS 15454 E100 card contains 24 ports, which multiplied by periods allows 96 rows in the table. An E1000 card contains 14 ports, which multiplied by four periods allows 56 table rows.
16.10.4.2 Row Creation in historyControlTable
The SetRequest PDU must be able to activate a historyControlTable row in one single-set operation. In order to do this, the PDU must contain all needed values and have a status variable value of 2 (createRequest). All OIDs in the SetRequest PDU should be type OID.0 type for entry creation.
To create a SetRequest PDU for the historyControlTable, the following values are required:
•
•
•
•
•
The historyControlBucketsRequested OID value is ignored because the number of buckets allowed for each sampling period, based upon the historyControlInterval value, is already fixed as listed in Table 16-9.
The historyControlInterval value cannot be changed from the four allowed choices. If you use another value, the SNMP agent selects the closest smaller time period from the set buckets. For example, if the set request specifies a 25-minute interval, this falls between the 15-minute (32 bucket) variable and the 60-minute (24 bucket) variable. The SNMP agent automatically selects the lower, closer value, which is 15 minutes, so it allows 32 buckets.
If the SetRequest PDU is valid, a historyControlTable row is created. If the row already exists, or if the SetRequest PDU values do not make sense or are insufficient, the SNMP agent does not create the row and returns an error code.
16.10.4.3 Get Requests and GetNext Requests
These PDUs are not restricted.
16.10.4.4 Row Deletion in historyControl Table
To delete a row from the table, the SetRequest PDU should contain a historyControlStatus value of 4 (invalid). A deleted row can be recreated.
16.10.5 Ethernet History RMON Group
The ONS 15454 implements the etherHistoryTable as defined in RFC 2819. The group is created within the bounds of the historyControlTable and does not deviate from the RFC in its design.
16.10.5.1 64-Bit etherHistoryHighCapacityTable
64-bit Ethernet history for the HC-RMON-MIB is implemented in the etherHistoryHighCapacityTable, which is an extension of the etherHistoryTable. The etherHistoryHighCapacityTable adds four columns for 64-bit performance monitoring data. These two tables have a one-to-one relationship. Adding or deleting a row in one table will effect the same change in the other.
16.10.6 Alarm RMON Group
The Alarm group consists of the alarmTable, which periodically compares sampled values with configured thresholds and raises an event if a threshold is crossed. This group requires the implementation of the event group, which follows this section.
16.10.6.1 Alarm Table
The NMS uses the alarmTable to determine and provision network performance alarmable thresholds.
16.10.6.2 Row Creation in alarmTable
To create a row in the alarmTable, the SetRequest PDU must be able to create the row in one single-set operation. All OIDs in the SetRequest PDU should be type OID.0 type for entry creation. The table has a maximum number of 256 rows.
To create a SetRequest PDU for the alarmTable, the following values are required:
•
•
•
•
•
•
If the SetRequest PDU is valid, a historyControlTable row is created. If the row already exists, or if the SetRequest PDU values do not make sense or are insufficient, the SNMP agent does not create the row and returns an error code.
In addition to the required values, the following restrictions must be met in the SetRequest PDU:
•
•
•
•
•
16.10.6.3 Get Requests and GetNext Requests
These PDUs are not restricted.
16.10.6.4 Row Deletion in alarmTable
To delete a row from the table, the SetRequest PDU should contain an alarmStatus value of 4 (invalid). A deleted row can be recreated. Entries in this table are preserved if the SNMP agent is restarted.
16.10.7 Event RMON Group
The Event group controls event generation and notification. It consists of two tables: the eventTable, which is a read-only list of events to be generated, and the logTable, which is a writable set of data describing a logged event. The ONS 15454 implements the logTable as specified in RFC 2819.
16.10.7.1 Event Table
The eventTable is read-only and unprovisionable. The table contains one row for rising alarms and another for falling ones. This table has the following restrictions:
•
•
•
•
16.10.7.2 Log Table
The logTable is implemented exactly as specified in RFC 2819. The logTable is based upon data that is locally cached in a controller card. If there is a controller card protection switch, the existing logTable is cleared and a new one is started on the newly active controller card. The table contains as many rows as provided by the alarm controller.
Posted: Wed Sep 5 06:48:58 PDT 2007
All contents are Copyright © 1992--2007 Cisco Systems, Inc. All rights reserved.
Important Notices and Privacy Statement.
