Cisco E-DI Concepts

Table Of Contents

Cisco E-DI Concepts

Overview

Cisco E-DI Features

Types of User

Security in Cisco E-DI

Role Based Access Control

Syslog Events

Device Authentication

Non Session Based Device Authentication

Session Based Device Authentication

Platform/OS Support

Communicating With Devices

Network Virtualization

Command Line Interface (CLI) and Prompt

Network Mode

Server Mode

Cisco E-DI Prompt

CLI

CLI Command Mode and Command Context

CLI Color Mode

Cisco E-DI admin Account

Cisco E-DI Root Login

Configuring Multiple Devices

Syntax Checking

Credential Sets

Groups

Static Groups

Dynamic Groups

Device Capabilities

Interface Groups

System Defined Groups

MyGroup Dynamic Group

Event Handling

Cisco E-DI Concepts

---

Cisco Enhanced Device Interface (Cisco E-DI) provides a comprehensive management interface for Cisco devices. See Figure 1-1. This chapter contains the following information:

•Overview

•Cisco E-DI Features

•Types of User

•Security in Cisco E-DI

•Platform/OS Support

•Communicating With Devices

•Network Virtualization

•Command Line Interface (CLI) and Prompt

•Configuring Multiple Devices

•Syntax Checking

•Credential Sets

•Groups

•Event Handling

Overview

Cisco E-DI provides interfaces for two main categories of users:

•The human user interacting with network devices through the command line interface (CLI).

•Management application programs interacting with network devices through a programmatic interface (PI).

Most of the Cisco devices natively provide a comprehensive CLI for a human user to handle all device level management functions. Cisco E-DI builds upon that capability providing value added functions to manage groups of devices conveniently, while keeping the new commands consistent with Cisco IOS CLI. Cisco E-DI provides an intuitive syntax validation of the commands, easy visual feedback on configurable and operational aspects of multiple devices in the network, and running CLI scripts on groups of devices.

While CLI scripts and macro commands can provide some programmatic support for managing large networks, the approach can still be cumbersome and unsuitable for comprehensive management of large networks. Management applications handling multi-vendor devices expect a standards based programmatic interface. Cisco E-DI provides an XML (eXtensible Markup Language) PI based on NETCONF configuration protocol standards. The supported data model is published through XSD (XML Schema Definitions) files.

Figure 1-1 Cisco E-DI

Management functions on Cisco devices can be classified as:

•Configuration, for example Conf

•Operational control, for example EXEC commands

•Operational data retrieval, for example Show commands

•Notifications (alarms and syslogs)

•Device troubleshooting and debugging

•Device software updates/upgrades

Functionality natively supported on a Cisco device is always available to the Cisco E-DI user.

Cisco E-DI primarily offers an enhanced interface for the following device level tasks:

•Configuring a device through XML PI and CLI

•Implementing EXEC commands on a device through XML PI and CLI

•Viewing a device file system through CLI

•Viewing device events through CLI

•Viewing and modifying device software through CLI

•Viewing device information and status through CLI

Cisco E-DI includes network virtualization for managing multiple devices. Network virtualization allows the user to dynamically group multiple devices into a single entity, and perform any of the tasks on all the selected devices.

A Cisco E-DI server manages a group of devices. However, if a user chooses to deploy multiple Cisco E-DI servers to manage the network, the partitioning of the network and which server will manage what partition of the network, will be the user's responsibility. Cisco E-DI is agnostic about another Cisco E-DI managing the network.

Cisco E-DI Features

Cisco E-DI includes the following features:

•XML programmatic interface

•CLI network virtualization

•Network management

•Session based device authentication

•Platform/OS support

•Configuration file management

•Basic inventory management

•Job scheduling

•Network troubleshooting and diagnostics

•Information synchronization

•Alarm/event handling

•Integral FTP server

•Server management

•Perl scripting capabilities

•Security management

•Color mode configuration

Types of User

There are four types of Cisco E-DI users:

•Administrator—Has full access and all privileges, and is considered to be the highest level of access.

•Network operator—Can perform any network related operations on Cisco E-DI; cannot modify the configuration or setup information. Multiple network operator accounts are possible, and can be defined by the administrator.

•Read-only user—Has read-only permissions, and cannot modify configurations.

•No access user—The default privilege over the domain's defined device group and the server. Privileges are to restrict any operation on the defined domain.

In addition, there is a repair user who can perform limited administration tasks like resetting the admin password, and recreating server database tables from the repair shell. See "Maintaining Cisco E-DI" for more details.

Security in Cisco E-DI

This section includes the following information:

•Role Based Access Control

•Syslog Events

•Device Authentication

–Non Session Based Device Authentication

–Session Based Device Authentication

Role Based Access Control

Users need to login to Cisco E-DI by specifying a username and password.

When a user attempts to login to Cisco E-DI in an SSH or Telnet session, the user is authenticated. On successful authentication, users are able to access the CLI.

Each user is associated with a domain group. The privileges for the domain group are specified when the domain group is created. For more information about domain groups and control, see "Domain Control" section.

Cisco E-DI provides role based access control at the device level, so that access to an NE is allowed or disallowed based on access privileges configured by the Cisco E-DI administrator.

To enable role based access control, any new user has to be associated with an existing domain group. Each module contains the task information for which it is responsible. Authorization information for each task is maintained in XML files for each privilege level. This information is loaded at the time the Cisco E-DI server is started.The command will only be implemented if the authorization check is successful. For example, a Cisco E-DI user can perform a management task on an NE only if:

•The user is in the FULL_CONTROL domain group

or

•The NE is one of the devices in the domain group, and the task the user is trying to perform, is permitted in the domain group

Syslog Events

All logins and configuration changes done in Cisco E-DI are published as Cisco E-DI syslog events and stored in the database. The events contain the name of the user who logged in successfully. Configuration change events contain the name of the user who made the changes.

External syslog receivers are able to receive the syslog events by subscribing for the events. Syslog events can be subscribed to using the Cisco E-DI server configuration command logging host <IP-ADDRESS>.

Device Authentication

Device authentication allows the administrator to choose between a centralized credential model (non session based device authentication) and a per user-session credential model (session based device authentication). Whichever mode is chosen is applied to all devices in the network.

During installation of Cisco E-DI, the administrator is prompted to choose between session based and non session based device authentication mode. Non session based device authentication is selected by default. A user can switch between session based and non session based device authentication mode at any time.

In Cisco E-DI, SNMP read-community and write-community credentials and enable password are shared, and are non session based irrespective of the device authentication mode. The administrator must specify the SNMP read community in the credential set. If required, the administrator can use domain grouping to restrict user access to devices. See Groups.

Non Session Based Device Authentication

Non session based device authentication can be used in an environment where there is no external AAA server.

If the administrator selects non session based device authentication during installation, Cisco E-DI prompts the administrator to set up the syslog auto subscription feature.

Cisco E-DI uses credential sets which are centralized (non session based) device credential stores. The administrator is required to pre-configure Cisco E-DI with the following device information in a credential set:

•SNMP Read Community

•SNMP Write Community (optional)

•Telnet/SSH login username

•Telnet/SSH password

•Enable password

•The transport type used for CLI sessions between Cisco E-DI and the device. (telnet or SSH)

Once these credentials are pre-configured for the managed devices, Cisco E-DI will use them automatically or on demand.

Session Based Device Authentication

When a network includes an external AAA server such as Cisco Secure Access Control Server, then non session based device authentication is unsuitable. For these networks, Cisco E-DI provides session based device authentication which requires a user to enter a login and password when managing devices. The device authentication login and password are valid for the entire duration of the user session, and are used for authenticating all the devices. The session login and password is not stored in Cisco E-DI.

Session based device authentication requires a user to set a login and password in the session in order to run the following commands:

•diag connectivity, and variations of this command

•diag device

•config setup, through the CLI or XML PI

•commit, in config setup mode,

•sync config fg|bg

•sync filesystem fg|bg

•inventory

•connect exec-mode

•exec-cmd <command>, through the CLI or XML PI.

•copy <from-device> <to-server>, includes the more command.

•copy <from-server> <to-device>. In this case, the destination filename is either deviceip:running-config or deviceip:startup-config.

•write [mem]

•reload device <ipaddress>

A user can overwrite the credentials stored in the session, in which case any new connections opened to devices from that point on will use the new credentials. The user must re-enter the session login and password if the Cisco E-DI session times out or is disconnected.

Passwords are destroyed from memory as soon as a user session terminates. A user can also explicitly delete passwords from memory by using the terminal no device-auth command.

When session based device authentication is enabled, the administrator can configure a central login and password in the credential set that will be used for background configuration and file system synchronizations.

Session based device authentication can be turned on or off during the Cisco E-DI installation. It can also be enabled or disabled at any time by the system administrator using the [no] device-auth session-based command.

If the administrator selects session based device authentication, Cisco E-DI automatically sets syslog auto-subscription to off. The user will be notified that each managed device must be configured to forward the syslog messages to Cisco E-DI either directly or through a syslog relay agent.

When session based device authentication is enabled, any login and password configured in the credential sets will be used only for background configuration and file system synchronizations.

Platform/OS Support

The actual devices, line-cards and OS releases supported by Cisco E-DI are determined and identified by the incremental device update (IDU) process and published on a regular basis. Refer to Release Notes for Cisco Enhanced Device Interface 2.0.1.

Communicating With Devices

Cisco E-DI uses SNMP, TFTP, and Telnet or SSH protocols to communicate with devices. Cisco E-DI supports SSH v1.5 and SSH v2. The administrator has the option of choosing Telnet or the required version of SSH. The user must specify a login, password, and an enable password for the chosen protocol. All device credentials, such as SNMP community strings and CLI passwords, are encrypted and stored in the startup configuration of the Cisco E-DI. See Credential Sets for more details.

Certain operations on a device can be destructive, for example, write erase which will erase the entire contents of flash on a device. Cisco E-DI provides a default list of forbidden commands, and administrators are able to modify the list. In Cisco E-DI, the administrator can define a list of commands which are not allowed to be implemented on any device. See Session Based Device Authentication.

Network Virtualization

Cisco E-DI includes the concept of network virtualization, where a network (a subnet, a network in a building, a group of devices) is seen as a single virtual device. For more information about groups, see Groups.

Command Line Interface (CLI) and Prompt

Cisco E-DI allows the user to interact with network devices through the command line interface (CLI). Cisco devices natively provide a comprehensive CLI for a user to handle all device level management functions.

There are three related concepts in Cisco E-DI:

•Command context

•Virtual File System (VFS)—Integrates Cisco E-DI server's file system and the managed device's file system into one directory structure, allowing the user to navigate through the file systems from a single console. See Figure 1-2.

•Command mode

Command context and VFS are related. The command context set commands, network and server, and VFS directory command cd change both the command context and the VFS directory path, and enable navigation within the Cisco E-DI main command. This behavior differs from a traditional operating system's shell. For example, in UNIX, there is only one command (cd) that changes the directory path. See Table 1-1 for examples.

Figure 1-2 Cisco E-DI Virtual File System

Network Mode

In Cisco E-DI setting the command context to network using the network command changes the working directory at the same time to /network. Cisco E-DI network configuration command mode is used for configuring devices on the network. This mode contains configuration sub-modes based on the specific device or devices types, and associated software version.

In this mode, operations apply on the devices in the network. The prompt signifies on what sub-set of devices the actions will be performed. In network mode there are two pre-defined sub-modes which in turn have multiple sub-modes. :

•Groups

•Devices

Server Mode

Setting the command context to server using the server command changes the working directory to /server. Cisco E-DI server configuration command mode is used for configuring the Cisco E-DI server.

In this mode all server related functions can be performed. For example, use server mode to see all the users and their rights, to configure credential sets and to implement other configuration commands.

Cisco E-DI Prompt

The Cisco E-DI run-time prompt includes the VFS path. This indicates the path which changed as a result of the command context command. It also shows the present directory as the user navigates through the directory structure to provide more context.

The Cisco E-DI prompt format is as follows:

User@Hostname[CommandContext:DirectoryPath]#

User@Hostname[CommandContext:DirectoryPath](CommandMode)#

The command context is shown as follows:

•SVR—Server

•NET—Network

•DEV—Device

•GRP—Group of devices

Table 1-1 shows examples of the Cisco E-DI prompt. CLI Command Mode and Command Context details the command context.

---

Note To avoid long pathnames in DirectoryPath, paths under /network/devices and /network/groups will have the /network/devices and /network/groups prefix replaced with ~ character. Others are to be shown in their full path.

---

Table 1-1 Cisco E-DI Prompt Examples 

Examples
Welcome to Cisco Management Switch (1.2) Copyright (C) 2005 Cisco System, Inc. All rights reserved. [Terminal vty4 Size 25x80] admin@hostname[SVR:/]#
admin@hostname[SVR:/]# cd server admin@hostname[SVR:/server]#
admin@hostname[SVR:/server]# server maintenance admin@hostname[SVR:/server](server-maint)#
admin@hostname[SVR:/server]# config t You are entering SERVER configuration mode. admin@hostname[SVR:/server](config)#
admin@hostname[SVR:/server]# network You are now in network view. Your present working directory: /network admin@hostname[NET:/network]#
admin@hostname[NET:/network]# cd devices admin@hostname[NET:/network/devices]#
admin@hostname[NET:/network/devices]# cd 172.16.0.0 admin@hostname[DEV:~/172.16.0.0]#
admin@hostname[DEV:~/172.16.0.0]# cd disk0: admin@hostname[DEV:~/172.16.0.0/disk0:]#
admin@hostname[DEV:~/172.16.0.0]# conf setup You are entering network config-setup mode Note The behavior of this command changes when session based device authentication is enabled. See Using Session Based Device Authentication for a full explanantion of the command behavior.
admin@hostname[DEV:~/172.16.0.0] (config-setup)# config t You are entering network configuration mode. Number of devices selected: 1 Selected device types: ---------------------- Device Type     No. of        Version                 Devices ----------------------------------------- Cisco7200          1          12.2(14)T2 admin@hostname[DEV:~/172.16.0.0](config)#
admin@hostname[NET:/network]# network group Routers You are now in network view. Your present working directory: /network/groups/Routers/ admin@hostname[GRP:~/Routers]#

CLI

Cisco E-DI NetCLI is the primary user interface to Cisco E-DI and is a CLI editor whose command editing facility is similar to that of Cisco IOS. To leverage the existing knowledge base in the Cisco user community, Cisco E-DI CLI is Cisco IOS CLI-like, Cisco E-DI NetCLI follows the same basic Cisco IOS CLI rules and behavior, for example, the concept of a no command to delete a configuration item.

The Cisco E-DI CLI includes the following features:

•CLI Command Mode and Command Context—Supports various CLI modes and context. Some of the examples of CLI modes are:

–Main Cisco E-DI command mode

–Cisco E-DI server configuration mode

–Cisco E-DI server maintenance mode

–Device and group configuration command mode

–Device and group tunneled EXEC command mode

Additionally, a pass-through command is provided to send a single unvalidated EXEC command.

•CLI Color Mode—Color is used for status summary in the CLI prompt and to indicate syntax validity by highlighting the entered command text.

•Filter/Pipe—All commands that produce text output to the screen supports | options to pipe the text output to e-mail, filtering criteria or redirection to a file as follows:

–append—Appends the data to a file.

–begin—Begins with the word that matches.

–email—Sends an email.

–exclude—Exclude lines that match a pattern.

–include—Include lines that match a pattern.

–save—Saves the data to a file.

•CLI User Interface:

–Supports creation of a minimum of 100 user accounts.

–Supports a minimum of 20 concurrent active user sessions.

CLI Command Mode and Command Context

As in Cisco IOS, the Cisco E-DI CLI command mode determines the set of commands available to the user at the prompt. However, in addition to command modes within the main Cisco E-DI command mode, the CLI also uses command context to determine the entity that the command is applied to, whether or not the command is applicable, and if it is applicable, how to interpret the command.

There are six main CLI command modes and the four command contexts within the main Cisco E-DI command mode. The Cisco E-DI prompt indicates the context (Server, Network, Group or Device) and path (Directory).

The six basic CLI modes are:

1. Main Cisco E-DI command mode—This is the main menu of the system containing main commands such as directory and file related commands, diagnostics commands, scripting commands, and commands to enter sub-modes. The main Cisco E-DI command mode also includes a subset of device and group EXEC commands that are syntax-validated and interpreted and implemented by Cisco E-DI. The outputs of these commands are generated by Cisco E-DI by interpreting the output generated by the device, not directly by the device itself.

2. Device and group configuration setup command mode—This mode is where all network related configurations are performed. The config-setup mode contains commands for entering into config mode for selected devices or combinations of devices to save, commit, schedule or discard configuration changes.

3. Device and group configuration command mode—Contains device and group configuration commands that are syntax-validated, but not interpreted or processed by Cisco E-DI. This mode is related to the configuration setup command mode. The commands are sent to the device as a group at the end of the configuration session.This mode contains configuration sub-modes based on the specific device or devices types, and its software version.

4. Cisco E-DI server configuration command mode—Contains commands for configuring the Cisco E-DI server.

5. Cisco E-DI server maintenance command mode—Contains commands for maintaining the Cisco E-DI server.

6. Device and group tunneled EXEC command mode—Device and group non-configuration device commands that are syntax-validated, but not interpreted or processed by Cisco E-DI. Commands are immediately sent to the device as soon as they are validated and the output, produced by the device, is shown to the user. An option to send a single non-syntax-validated EXEC command from the main Cisco E-DI command mode is also provided.

Figure 1-3 shows the hierarchy and relationship of the 6 basic CLI modes.

Figure 1-3 Hierarchy and Relationship of the CLI Modes

The four command contexts within the Cisco E-DI main command mode are:

1. Server (SVR)—Context is set to server when the working directory is /, /server and its subdirectories, and /users and its subdirectories.

2. Network (NET)—Context is set to network when the working directory is /network, /network/devices, and /network/groups.

3. Device (DEV)—Context is set to device when the working directory is /network/devices/<device id>.

4. Group (GRP)—Context is set to group when the working directory is /network/devices/<group id>.

The concept of a group is an integral part of Cisco E-DI. For more information, see Grouping. Command context is associated with the Cisco E-DI VFS directory path.

Navigation between command contexts and VFS directories within the Cisco E-DI main command mode is supported by both the network and server command, and by the directory change command cd.

CLI Color Mode

Cisco E-DI CLI color mode enhances CLI usability by coloring the display headings and CLI mode in the prompt. The NE's alarm aggregate status is indicated in the CLI prompt. The color of the prompt indicates the highest alarm severity found in the devices within the scope of the CLI mode, as follows:

•Red—If any one of the devices has a P1 alarm.

•Yellow—If all the devices have P2 and lower alarms.

•Green—When the devices have no alarms on them.

Cisco E-DI color mode has been tested on the following terminal types:

•Putty (open source client for SSH and Telnet)

•Token2 (Open source Telnet Client)

•Windows DOS Telnet application

•Windows Hyperterminal

Additionally, each command typed by the user instantly gets color highlighted to indicate the validity of the command. For example, if the user enters the word hostne for the hostname command, the text will be highlighted blue till the word hostn is entered, but as soon as e is typed, the word hostne will be highlighted red to indicate that there is no matching command for that word.

Cisco E-DI admin Account

Cisco E-DI provides a pre-defined admin account. The name of the account may not be changed by any user, but the password can be changed. Users with FULL_CONTROL access are considered to be Cisco E-DI administrators. Any Cisco E-DI administrator can add more administrators or other user accounts in Cisco E-DI using the CLI commands.

Cisco E-DI Root Login

Access to the root login is fully restricted since the root login and password is disabled.

An incremental root login package that enables root login is available as a patch for debugging and serviceability purposes. On applying this patch, a user can gain root privileges. The purpose of the root login is to allow TAC, or the user with instructions from TAC, to troubleshoot basic Linux functions.

---

Note Refer to the Cisco Enhanced Device Interface 2.0.1 Quick Start Guide for details about how to obtain patches.

---

Configuring Multiple Devices

Through the network virtualization feature, Cisco E-DI provides an ability to configure multiple devices with one set of commands. When a user selects multiple devices of different types and/or different OS versions, Cisco E-DI automatically determines the least common denominator set of commands and presents them to the user in typical Cisco IOS-like fashion. If the color mode is enabled, then the command is highlighted with an appropriate color to give the user an instant feedback on the validity of a command with the given device/OS selection. See CLI Color Mode for more information. The user can then press the key combination Ctrl-G to view the detailed mapping of a command for any given device type/OS version.

When configuring multiple devices, Cisco E-DI provides interface macros, for example all-fastethernet, all-gigabitethernet. When configuring a single device, Cisco E-DI provides an additional macro for selecting a single interface.

Syntax Checking

Cisco E-DI maintains the CLI knowledge base for every device family/OS version that it supports. With this knowledge base, Cisco E-DI can perform automatic syntax checking on all user input. Cisco E-DI also internally uses the syntax checking feature to intelligently identify changes between two configurations.

Cisco E-DI also allows the user to select syntax checking of additional options such as OS version, and OS type.

Credential Sets

Device credentials like login, password, and SNMP community string settings are required for communication with a device. Cisco E-DI combines these credentials into a credential set which specifies the necessary information for Cisco E-DI to communicate to the device. It is assigned to a device when the device is managed.

---

Note The behavior of the login and password commands changes when session based device authentication is enabled. See Using Session Based Device Authentication for a full explanantion of the command behavior.

---

Credential sets have the following attributes:

•Community strings—SNMP read and write community strings.

•Login credentials—Username, password and enable password to Telnet to a device.

•Console server—Terminal server's IP address and port information for devices accessible through the terminal server.

•Transport type—Transport options are either Telnet, or SSH v1.5, or SSH v2. This selection is pre-determined by the administrator at the time of managing the device. Choice of communication protocol will not be available to each session. The following ciphers are supported for SSH:

–3des—Triple-DES cipher. This is the default cipher type for SSH.

–aes_128—AES cipher (128 bit)

–aes_192—AES cipher (192 bit)

–aes_256—AES cipher (256 bit)

–arcfour—Arc Four cipher (SSH v2 only)

–blowfish—Blow-fish cipher

–des—DES cipher (SSH v1.5 only)

–twofish—Two-fish cipher

The following modes can be configured for SSH v2:

–cbc

–cfb

–ctr

–ecb

–ofb

•Encryption level—Encryption level of the password. Passwords will always be encrypted when displayed. Users have the option to select an encryption level when specifying a password.

•Based-on—Allows the current credential set to inherit attributes from the credential set that follows this option, except for the defined attributes in the current credential set and the terminal server attribute.

---

Note A default credential set exists with SNMP read community string set to public. This set will be used if no other credential set is assigned for a device.

---

Cisco E-DI uses the attributes defined in the credential set to login to the device, and to perform SNMP operations. A credential set can be assigned to a single device or multiple devices. If there is no credential set assigned to a device, default credential set will be used.

Credential sets can also be used for troubleshooting, where the user specifies the credential set to be used for trying connectivity to the device.

Groups

Cisco E-DI provides the option to create groups. This can be used to manage groups of devices conveniently. There are the following types of groups:

•Device grouping—Provide context for the Cisco E-DI CLI operations. There are two types of NE groups:

–Static Groups—Selecting one or more NE or group.

–Dynamic Groups—Using a grouping criteria.

•Interface Groups—Sets of static system-defined groups that combine multiple network interfaces into a single interface which may be used by the user for configuring several interfaces at once. For example, an interface group could be all-Ethernet, or all-fast-Ethernet.

•System Defined Groups—These groups are pre-defined by the system. They cannot be modified or deleted. Following system groups exist now: AccessPoints, CiscoAP1100, CiscoAP1200, CiscoAP350IOS, Switches, Routers, Firewalls, IDS, CompleteNetwork, and Unknown.

•MyGroup Dynamic Group—A group created by the user that can contain any managed device based on the context the user chooses. The selection of devices in this group is not persistent, and is lost on exit from the group.

Once the NEs are grouped, then the grouped NEs can be placed in a domain group, that is, the administrative domain.

Groups can also be nested within a group.

Groups are fundamental to the concept of network virtualization (see Network Virtualization) where users can dynamically group multiple devices into a single entity, and perform any of the tasks on all the selected devices.

When network mode is selected, operations apply on the devices in the network. The prompt signifies on what sub-set of devices the actions will be performed. In network mode there are two sub-modes which in turn have multiple sub-modes. There are two pre-defined sub-modes under the network mode; groups and devices.

Static Groups

Static groups contain a statically defined set of devices, or other groups which can be static, dynamic or system defined, creating a nested group.

Devices are added to these groups statically. A user can add another static or dynamic group to a static group, to create a nested group. The devices contained in the nested groups are the list of all devices contained in all of the groups included in the nested groups with redundant devices listed only once.

Only managed devices can be added to a static group. Devices can be added to or deleted from a static group.

Dynamic Groups

Dynamic groups are rule-based. Devices are grouped together based on user-defined rules. The list of devices is dynamically computed based on user-defined rules. Whenever a device is managed, it becomes a part of a dynamic group if it satisfies the rules specified for that group. Devices cannot be removed from a dynamic group. Devices can be prevented from being added to a group if exclude options are included in the rules for that group. Rule features include:

•Include or exclude capabilities. See Device Capabilities.

•Range of IP addresses

•Device name pattern

•Device type

•Device family

Device Capabilities

Device capability is a unique name that identifies certain capabilities that a device supports. For example: cdp-mib-supported. Capabilities are used by Cisco E-DI to determine which Cisco E-DI functionalities are applicable to the device. Table 1-2 lists the capabilities supported by Cisco E-DI:

Table 1-2 Device Capabilities 

Name	Device capability
bridge-mib-supported	Device capability bridge-mib-supported
cdp-enabled	Device capability cdp-enabled
cdp-mib-supported	Device capability cdp-mib-supported
dot11-ap	Device capability dot11-ap
dot11-infrastructure-client-mode	Device capability dot11-infrastructure-client-mode
dot11a-radio	Device capability dot11a-radio
dot11b-radio	Device capability dot11b-radio
dot11g-radio	Device capability dot11g-radio
edi-server	Device capability edi-server
entity-mib-supported	Device capability entity-mib-supported
firewall	Device capability firewall
flash-mib-supported	Device capability flash-mib-supported
generic-bridge	Device capability generic-bridge
generic-host	Device capability generic-host
ids	Device capability ids
ios-style-commands	Device capability ios-style-commands
l2-switch	Device capability l2-switch
l3-router	Device capability l3-router
nms-platform	Device capability nms-platform
old-cisco-chassis-mib-supported	Device capability old-cisco-chassis-mib-supported
os-type-catos	Device capability os-type-catos
os-type-ios	Device capability os-type-ios
os-type-pixos	Device capability os-type-pixos
radio-monitor-mode	Device capability radio-monitor-mode
stack-mib-supported	Device capability stack-mib-supported
stp-supported	Device capability stp-supported
sylog-source	Device capability sylog-source
tftp-client	Device capability tftp-client
tftp-server	Device capability tftp-server
unknown-device-type	Device capability unknown-device-type
vpn	Device capability vpn
vtp-mib-supported	Device capability vtp-mib-supported
vtp-supported	Device capability vtp-supported

Interface Groups

Interface group is a static system-defined groups used within a device context. A device context could cover single or multiple devices.

Interface groups allow user to configure multiple interfaces with one set of commands. Interface groups can be used with multiple devices or a single device. The interface grouping feature is only available through the CLI.

Table 1-3 lists the supported interface groups.

Table 1-3 Interface Groups 

Name	Description
all	Interface group all
all-atm	Interface group all-atm
all-bvi	Interface group all-bvi
all-dot11	Interface group all-dot11
all-dot11a	Interface group all-dot11a
all-dot11b	Interface group all-dot11b
all-dot11g	Interface group all-dot11g
all-ethernet	Interface group all-ethernet
all-fast-ethernet	Interface group all-fast-ethernet
all-ge-wan	Interface group all-ge-wan
all-gigabit-ethernet	Interface group all-gigabit-ethernet
all-loopback	Interface group all-loopback
all-pos	Interface group all-pos
all-serial	Interface group all-serial
all-vlan	Interface group all-vlan

System Defined Groups

There are several pre-defined dynamic groups known as system defined groups, as follows:

•AccessPoints—All Cisco access points in the network

•CiscoAP1100—All Cisco AP1100 devices

•CiscoAP1200—All Cisco AP1200 devices

•CiscoAP350IOS—All CiscoAP350IOS devices

•CompleteNetwork—All devices currently managed by Cisco E-DI

•FireWalls—All Cisco firewalls

•IDS—All Cisco IDS systems

•MyGroup—List of devices currently in user context (for example, if user selects one or more devices with command 'network <ip-addr1> <ip-adddr2>' then both these devices are kept in MyGroup. The contents of MyGroup are specific to user session.

•Routers—All L3 routers

•Switches—All L2 switches

•Unknown—All devices whose type is unknown to Cisco E-DI

MyGroup Dynamic Group

MyGroup is an ad-hoc dynamic group which is created by the user, and is lost when the user exits the session. MyGroup can be any combination of devices and existing groups.

For example, MyGroup would be created when a command similar to the following is entered:

[SVR:/server] (config)# network 172.16.0.202   172.16.0.200 172.16.0.204

You can enter [SVR:/server]# show devices group to show the devices in the group.

Event Handling

Cisco E-DI uses events received from devices, and data collected during polling and inventory, to maintain an accurate representation of the state of the devices and the network. This data provides the user and higher level applications with an up-to-date view of the health of the device, and alerts the user to configuration changes that are likely to fail. All events received from the network are automatically archived in the database.

Cisco E-DI receives syslog events directly from NEs, or from syslog servers in the network. The syslog event uses the DNS name and IPv4 address to identify the NE. The source IP address can be any interface on a managed device.

If an event is received through two different paths, for example, directly from the NE, and also from a relay agent, Cisco E-DI archives both the events as if they are two different events. As Cisco  E-DI performs event collating, duplicate events typically do not cause replicate synchronization actions.

Syslog events are archived in the database according to the timestamp when the event was received by Cisco E-DI.

Cisco E-DI supports the Kiwi Syslog Server.

When creating an action in the Kiwi Syslog Server to forward syslogs to Cisco E-DI, select the option Retain the original source address of the message. If this option is not selected, syslogs will be forwarded without the device IP address, and Cisco E-DI will drop the forwarded syslog messages.

Any event received from an unmanaged entity or from unknown relay agent is dropped.

Statistics of the number of relay events received, and events dropped are maintained. You can view these statistics, enter show server stats.