Cisco SCE8000 10GBE Software Configuration Guide, Release 3.7.x
Configuring the Management Interface and Security

Table Of Contents

Configuring the Management Interface and Security

Introduction

Management Interface and Security

Configuring the Management Ports

Entering the Management Interface Configuration Mode

Configuring the Management Port Physical Parameters

Setting the IP Address and Subnet Mask of the Management Interface

Configuring the Management Interface Speed and Duplex Parameters

Specifying the Active Management Port

Management Interface Redundancy

Configuring the Management Ports for Redundancy

Configuring the Fail-Over Mode

Monitoring the Management Interface

Configuring Management Interface VLANs

Monitoring Management VLANs

Mapping of VLAN ID to Virtual Gi ID

Gi AVP

Configuring VLAN ID Mapping to Virtual Gi ID

Monitoring Mapping of VLAN ID to Virtual ID

TACACS+ Authentication, Authorization, and Accounting

Information about TACACS+ Authentication, Authorization, and Accounting

Login Authentication

Accounting

Privilege Level Authorization

Command Level Authorization

General AAA Fallback and Recovery Mechanism

About Configuring TACACS+

Configuring the SCE Platform TACACS+ Client

Adding a New TACACS+ Server Host

Removing a TACACS+ Server Host

Configuring the Global Default Key

Configuring the Global Default Timeout

Managing the User Database

Adding a New User to the Local Database

Defining the User Privilege Level

Adding a New User with Privilege Level and Password

Deleting a User

Configuring AAA Login Authentication

Configuring Maximum Login Attempts

Configuring the AAA Login Authentication Methods

Configuring AAA Privilege Level Authorization Methods

Configuring AAA Command Level Authorization Methods

Configuring AAA Accounting

Monitoring TACACS+

Displaying Statistics for TACACS+ Servers

Displaying Statistics, Keys and Timeouts for TACACS+ Servers

Monitoring TACACS+ Users

Configuring Access Control Lists (ACLs)

Adding Entries to an ACL

Removing an ACL

Defining a Global ACL

Managing the Telnet Interface

Preventing Telnet Access

Assigning an ACL to the Telnet Interface

Removing the ACL Assignment from the Telnet Interface

Configuring the Telnet Timeout

Configuring the SSH Server

The SSH Server

Key Management

Managing the SSH Server

Generating a Set of SSH Keys

Enabling the SSH Server

Disabling the SSH Server

Running Only SSHv2

Assigning an ACL to the SSH Server

Removing the ACL Assignment from the SSH Server

Deleting the Existing SSH Keys

Monitoring the Status of the SSH Server

Configuring and Managing the SNMP Interface

About the SNMP Interface

SNMP Protocol

Security Considerations

About CLI

About MIBs

Configuration via SNMP

Enabling the SNMP Interface

How to Enable the SNMP Interface

How to Disable the SNMP Interface

Configuring SNMP Community Strings

Defining a Community String

Removing a Community String

Displaying the Configured Community Strings

Configuring SNMP Notifications

Defining SNMP Hosts

Configuring SNMP Traps


Configuring the Management Interface and Security


Revised: September 27, 2012, OL-24150-06

Introduction

This chapter describes how to configure the physical management interfaces (ports) as well as the various management interface applications, such as SNMP, SSH, and TACACS+. It also explains how to configure users, passwords, IP configuration, clock and time zone, and domain name settings.

Management Interface and Security

Configuring the Management Ports

Configuring Management Interface VLANs

Mapping of VLAN ID to Virtual Gi ID

TACACS+ Authentication, Authorization, and Accounting

Configuring Access Control Lists (ACLs)

Managing the Telnet Interface

Configuring the SSH Server

Configuring and Managing the SNMP Interface

Management Interface and Security

The SCE8000 platform is equipped with two RJ-45 management ports (Port1 and Port2 on the SCE8000-SCM-E module in slot 1). These ports provide access from a remote management console to the SCE platform via a LAN.

The two management ports support management interface redundancy, providing the possibility for a backup management link.


Note The second management port is reflected in all objects related to it in the SNMP interface.



Note Cisco SCE 8000 does not support IPv6 addresses or connectivity on the management interfaces.


Perform the following tasks to configure the management interface and management interface security:

Configure the management port:

Physical parameters

Specify active port (if not redundant installation)

Redundancy (if redundant installation)

Configure management interface security

Configure the permitted and not-permitted IP addresses

Configuring the Management Ports

Entering the Management Interface Configuration Mode

Configuring the Management Port Physical Parameters

Management Interface Redundancy

Monitoring the Management Interface

Perform the following tasks to configure the management ports:

Configure the IP address and subnet mask (only one IP address for the management interface, not one IP address per port).

Configure physical parameters:

Duplex

Speed

active port (optional)

Configure redundant management interface behavior (optional):

Fail-over mode


Step 1 Cable the desired management port, connecting it to the remote management console via the LAN. If connecting both management ports for redundancy, connect the to the LAN using a switch.

Step 2 Configure the management port physical parameters. (See Configuring the Management Port Physical Parameters.)

Step 3 (Optional) Configure the system with management interface redundancy. (see Management Interface Redundancy.)


Entering the Management Interface Configuration Mode

When entering Management Interface Configuration Mode, you must indicate the number of the management port to be configured:

0/1 — Port1

0/2 — Port2

The following Management Interface commands are applied only to the port specified when entering Management Interface Configuration Mode. Therefore, each port must be configured separately:

speed

duplex

The following Management Interface commands are applied to both management ports, regardless of which port had been specified when entering Management Interface Configuration Mode. Therefore, both ports are configured with one command:

ip address

auto-fail-over

The GBE management interface is configured as follows:

mode: MNG Interface configuration mode

interface designation: 0/1 or 0/2


Step 1 Enter configure and press Enter.

Enters Global Configuration mode.

The command prompt changes to SCE(config)#.

Step 2 Enter interface mng (0/1 | 0/2) and press Enter.

Enters Management Interface Configuration mode.

The command prompt changes to SCE(config if)#


Configuring the Management Port Physical Parameters

This interface has a transmission rate of 10, 100, or 1000 Mbps and is used for management operations and for transmitting RDRs, which are the output of traffic analysis and management operations.

Setting the IP Address and Subnet Mask of the Management Interface

Configuring the Management Interface Speed and Duplex Parameters

Specifying the Active Management Port

Setting the IP Address and Subnet Mask of the Management Interface

You must define the IP address of the management interface.

When both management ports are connected, providing a redundant management port, this IP address always acts as a virtual IP address for the currently active management port, regardless of which port is the active port.

Options

The following options are available:

IP address — The IP address of the management interface.

If both management ports are connected, so that a backup management link is available, this IP address will be act as a virtual IP address for the currently active management port, regardless of which physical port is currently active.

The following IP addresses are used internally by the SCE8000 platform and cannot be assigned to the management interface:

192.168.207.241 to 192.168.207.255

92.168.207.145 to 192.168.207.159

subnet mask — subnet mask of the management interface.


Step 1 On a physically connected local console, access the interface configuration mode for either management interface. The specified IP address is configured for both interfaces.

From the SCE(config)# prompt, type interface Mng (0/1 | 0/2) and press Enter.

Step 2 From the SCE(config if)# prompt, type ip address ip-address subnet-mask and press Enter.

The command might fail if there is a routing table entry that is not part of the new subnet defined by the new IP address and subnet mask.


Caution Changing the IP address of the management interface via telnet will result in loss of the telnet connection and inability to reconnect with the interface.


Note After changing the IP address, you must reload the SCE platform so that the change will take effect properly in all internal and external components of the SCE platform. (See Rebooting and Shutting Down the Cisco SCE Platform.)



Setting the IP Address and Subnet Mask of the Management Interface: Example

The following example shows how to set the IP address of the SCE platform to 10.1.1.1 and the subnet mask to 255.255.0.0.

SCE#config
SCE(config)#interface mng 0/1
SCE(config if)#ip address 10.1.1.1 255.255.0.0

Configuring the Management Interface Speed and Duplex Parameters

This section presents sample procedures that describe how to configure the speed and the duplex of the management interface.

Both these parameters must be configured separately for each port.

Interface State Relationship to Speed and Duplex

How to Configure the Speed of the Management Interface

How to Configure the Duplex Operation of the Management Interface

Interface State Relationship to Speed and Duplex

Table 5-1 summarizes the relationship between the interface state and speed and duplex.


Note It is not possible to set one parameter to "Auto" and to specify the other parameter. If either speed or duplex is set to "Auto", then both parameters will behave as if set to "Auto".


Table 5-1 Interface State Relationship to Speed and Duplex 

Speed
Duplex
Actual GBE Interface State

Auto

Auto

Auto negotiation

Auto

Full/Half

Auto negotiation

10/100/1000

Auto

Auto negotiation

10

Full

10 Mbps and full duplex

10

Half

10 Mbps and half duplex

100

Full

100 Mbps and full duplex

100

Half

100 Mbps and half duplex


How to Configure the Speed of the Management Interface

Options

The following options are available:

speed — speed in Mbps of the currently selected management port (0/1 or 0/2):

10

100

auto (default) — auto-negotiation (do not force speed on the link)

If the duplex parameter is configured to auto, changing the speed parameter has no effect.


Step 1 Access the interface configuration mode for the management interface you want to configure.

From the SCE(config)# prompt, type interface Mng (0/1 | 0/2) and press Enter.

Step 2 From the SCE(config if)# prompt, type speed (10|100|auto) and press Enter.

Specify the desired speed option.


Configuring the Speed of the Management Interface: Example

The following example shows how to use this command to configure the Management port to 100 Mbps speed.

SCE#config
SCE(config)#interface mng 0/1
SCE(config if)#speed 100

How to Configure the Duplex Operation of the Management Interface

Options

The following options are available:

duplex — duplex operation of the management port (0/1 or 0/2):

full

half

auto (default) — auto-negotiation (do not force duplex on the link)

If the speed parameter is configured to auto, changing the duplex parameter has no effect.


Step 1 Access the interface configuration mode for the management interface you want to configure.

From the SCE(config)# prompt, type interface Mng (0/1 | 0/2) and press Enter.

Step 2 From the SCE(config if)# prompt, type duplex auto|full|half and press Enter.

Specify the desired duplex option.


Configuring the Duplex Operation of the Management Interface: Example

The following example shows how to use this command to configure the management port to half duplex mode.

SCE#config
SCE(config)#interface mng 0/2
SCE(config if)#duplex half

Specifying the Active Management Port

This command explicitly specifies which management port is currently active. Its use varies slightly, depending on whether the management interface is configured as a redundant interface (auto fail-over enabled) or not (auto fail-over disabled).

auto fail-over enabled (automatic mode) — the specified port becomes the currently active port, in effect forcing a fail-over action even if a failure has not occurred.

auto fail-over disabled (manual mode) — the specified port should correspond to the cabled Mng port, which is the only functional port and therefore must be and remain the active management port.


Step 1 Access the interface configuration mode for the management interface you want to configure as the active management port.

From the SCE(config)# prompt, type interface Mng (0/1 | 0/2) and press Enter.

Step 2 Type active-port and press Enter


Specifying the Active Management Port: Example

The following example shows how to use this command to configure Mng port 2 as the currently active management port.

SCE#config
SCE(config)#interface mng 0/2
SCE(config if)#active-port

Management Interface Redundancy

Configuring the Management Ports for Redundancy

Configuring the Fail-Over Mode

The SCE platform contains two RJ-45 management ports. The two management ports provide the possibility for a redundant management interface, thus ensuring management access to the SCE platform even if there is a failure in one of the management links. If a failure is detected in the active management link, the standby port automatically becomes the new active management port.

Note that both ports must be connected to the management console via a switch. In this way, the IP address of the MNG port is always the same, regardless of which physical port is currently active.

Important information:

Only one port is active at any time.

The same virtual IP address and MAC address are assigned to both ports.

Default:

Port 1 = active

Port 2 = standby

The standby port sends no packets to the network and packets from the network are discarded.

When a problem in the active port is encountered, the standby port automatically becomes the new active port.

Link problem, with switch to standby MNG port, is declared after the link is down for 300 msec.

Service does not revert to the default active port if/when that link recovers. The currently active MNG port remains active until link failure causes a switch to the other MNG port.

Configuring the Management Ports for Redundancy


Step 1 Cable both management ports (Mng 1 and Mng 2), connecting them both to the remote management console via the LAN and via a switch.

Using the switch ensures that the IP address of the MNG port is always the same, regardless of which physical port is currently active

Step 2 Configure the automatic fail-over mode.

See Configuring the Fail-Over Mode.

Step 3 Configure the IP address for the management interface.

The same IP address will always be assigned to the active management port, regardless of which physical port is currently active.

See Setting the IP Address and Subnet Mask of the Management Interface.

Step 4 Configure the speed and duplex for both management ports.

See Configuring the Management Interface Speed and Duplex Parameters.


Configuring the Fail-Over Mode

Options

How to Enable the Automatic Fail-Over Mode

How to Disable the Automatic Fail-Over Mode

Use the following command to enable automatic fail-over. The automatic mode must be enabled to support management interface redundancy. This mode automatically switches to the backup management link when a failure is detected in the currently active management link.

This parameter can be configured when in management interface configuration mode for either management port, and is applied to both ports with one command.

Options

The following options are available:

auto/ no auto — Enable or disable automatic fail-over switching mode

Default — auto (automatic mode)

How to Enable the Automatic Fail-Over Mode

From the SCE(config if)# prompt, type:

Command
Purpose

auto-fail-over

Enables automatic fail-over mode.


How to Disable the Automatic Fail-Over Mode

From the SCE(config if)# prompt, type:

Command
Purpose

no auto-fail-over

Disables automatic fail-over mode.


Monitoring the Management Interface

Use this command to display the following information for the management interface.

speed

duplex

IP address

auto-fail-over configuration

From the SCE# prompt, type:

Command
Purpose

show mng interface Mng (0/1 | 0/2) [speed | duplex | |ip address | auto-fail-over]

Displays the specified GBE management interface configuration for the specified interface. If no option is specified, all management interface information is displayed for the specified interface.


Configuring Management Interface VLANs

The SCE management network interface is used for various management services such as:

Accessing the SCE shell through Telnet or SSH.

SNMP

Management interface VLANS provide a way to distinguish between these management services (Telnet, SSH, SNMP) by using separate VLANs carried over same physical port (see Figure 5-1).

Figure 5-1 Management Interface VLANS

There are two steps in configuring management VLANs:

1. Create the VLAN and assign an IP address (mng-vlan command).

2. Assign the VLAN to a management or control service. Use one of the following commands:

ip ssh mng-vlan

vty mng-vlan

snmp-server mng-vlan

When a new management VLAN is configured, the device adds a new routing entry in the routing table. So, each configured IP address should be routable from the Cisco SCE.

Here is an example of a sample configuration.

The following are some of the valid IP entries:

10.10.10.10/255.255.255.192 - VLAN ID 100

10.10.20.10/255.255.255.192 - VLAN ID 120

10.10.30.10/255.255.255.192 - VLAN ID 150

If the IP addresses listed above are added, the device creates the following entries in the routing table:

Destination Gateway Genmask Flags Metric Ref Use Iface

10.10.20.0 * 255.255.255.192 U 0 0 0 bond0.120

10.10.30.0 * 255.255.255.192 U 0 0 0 bond0.150

10.10.10.0 * 255.255.255.192 U 0 0 0 bond0.100

The following diagram provides another view of the configured management VLAN:

SUMMARY STEPS

1. enable

2. configure

3. mng-vlan vlan-id address ip-address mask mask

4. <service> mng-vlan vlan-id

DETAILED STEPS

 
Command
Purpose

Step 1 

enable

Example:

SCE> enable

Enables privileged EXEC mode. Enter your password when prompted.

Step 2 

configure

Example:

SCE#> configure

Enters global configuration mode.

Step 3 

mng-vlan vlan-id address ip-address mask mask

Example:

SCE(config)#> mng-vlan 10 address 10.10.10.20 mask 
255.255.255.0

Creates the VLAN and assigns the IP address to the VLAN.

Range for VLAN tag: 1-4094

Step 4 

<service> mng-vlan vlan-id

Example:

SCE(config)#> snmp-server mng-vlan 10

Assigns the VLAN to the specified service.

Service options are:

ip ssh

vty

snmp-server

Note The snmp-server mng-vlan command, in either the positive or negative form, restarts the SNMP process in order for the changes to take effect. This generates a cold-start trap.

Monitoring Management VLANs

To monitor managment VLANs, use one or more of the following commands.

These commands are in viewer mode.

Command
Purpose

show mng-vlan [vlan-id | all]

Displays the IP configuration and configured management service for the specified management VLAN.

show vlan vlan-id service-bind

Displays the management service assigned to the specified management VLAN.

show vlan vlan-id statistics

Displays the traffic statistics for the specified VLAN.

show vty mng-vlan

Displays the management interface VLAN configured for Telnet services.

show ip ssh mng-vlan

Displays the management interface VLAN configured for SSH services.


Mapping of VLAN ID to Virtual Gi ID

The mapping of VLAN ID to virtual Gi ID feature enables SCE to map the VLAN ID retrieved from the subscriber traffic to a virtual Gi ID; thus, allowing the PCRF to fetch the policy corresponding to the VLAN ID and IP address, and send it to SCE. The physical VLAN ID received from the subscriber side traffic is in the range of 1-4094. This range (1-4094) is mapped to a static virtual ID that is of the range 1-255, and is used by the PCRF server to fetch the policy.

The mapped virtual ID is sent to the PCRF server as part of the CCR-I request. The corresponding policy is sent back to the SCE as part of the CCA answer.

Figure 5-2 Mapping of VLAN ID to Virtual ID

Gi AVP

Table 5-2 defines the Gi AVP that carries the mapped Virtual Gi ID as part of a CCR-I request:

Table 5-2 Gi Interface AVP

AVP Name
AVP Code
Value Type
Used In

TMO-Virtual-Gi-ID

120

Uint32

CCR-I

CCA

RAR

RAA

 

O (optional)

Subscriber-ID

443

             

Subscription-ID-Type

450

             

Subscription-ID-Data

444


The optional flag (O) is set for TMO-Virtual-Gi-ID and is used in CCR-I. The value type (unsigned integer 32) carries the value of the mapped virtual-ID.

Configuring VLAN ID Mapping to Virtual Gi ID

This section contains the information and instructions to configure and monitor the feature.

Following are the steps for configuring the mapping of VLAN ID to a virtual Gi interface ID.

1. Configure the VLAN ID mapping to a virtual Gi ID.

2. Configure the Gx server.

3. Enable the virtual Gi mode in the interface configuration mode.

4. Enable VLAN symmetric classification.

SUMMARY STEPS

1. enable

2. configure

3. diameter gx virtual-gi vlan-id mapping

4. diameter peer

5. interface linecard 0

6. subscriber virtual-gi-mode

7. VLAN symmetric classify

8. show interface linecard 0 subscriber virtual-gi-mode

9. show diameter gx virtual-gi all

10. show running-config | include subscriber

11. show running-config | include diameter

DETAILED STEPS

 
Command
Purpose

Step 1 

enable

Example:

SCE> enable

Enables privileged EXEC mode. Enter your password when prompted.

Step 2 

configure

Example:

SCE#> configure

Enables the global configuration mode.

Step 3 

diameter gx virtual-gi vlan-id vlan-id mapping value1

Example:

SCE(config)#> diameter Gx virtual-gi vlan-id 2 
mapping 3

Configures the VLAN-ID to virtual-Gi-ID mapping in the global configuration mode.

vlan-id—Specifies the vlan-id to be mapped.

value1—Specifies the virtual Gi ID.

Step 4 

diameter peer name peer-host ip-address [port <port#>]

Example:

SCE(config)#> diameter peer GX peer-host 
10.78.241.155 port 3868

Configures the Gx server (PCRF server) in the global configuration mode.

name—Name to be assigned to the entry in the peer table.

ip-address—IP address of the host.

A peer is defined by an URI. This means that the same IP can not be used on different ports to distinguish between two servers except when a DNS is used.

port#—Port number used.

Step 5 

interface linecard 0

Example:

SCE(config)#> interface linecard 0

Enables the interface configuration mode.

Step 6 

subscriber virtual-gi-mode

Example:

SCE(config-if)#> subscriber virtual-gi-mode

Enables the virtual Gi mode.

Step 7 

VLAN symmetric classify

Example:

SCE(config-if)#> VLAN symmetric classify

Enables VLAN symmetric classification in the interface configuration mode. This command enables VLAN-ID to be retieved from the subscriber traffic.

Step 8 

show interface linecard 0 subscriber virtual-gi-mode

Example:

SCE(config)#> show interface linecard 0 subscriber 
virtual-gi-mode

Enter this command in the global configuration mode.

Displays the status of the virtual Gi mode.

Step 9 

show diameter gx virtual-gi all

Example:

SCE(config)#> show diameter gx virtual-gi all

Displays the virtual Gi mapping table having all the existing VLAN ID to virtual Gi ID mapping.

Step 10 

show running-config | include subscriber

Example:

SCE(config)#> show running-config | include 
subscriber

Displays the status of virtual Gi.

Step 11 

show running-config | include diameter

Displays diameter related configurations made in the system.

Monitoring Mapping of VLAN ID to Virtual ID

To monitor mapping of VLAN ID to Virtual ID, use one or more of the following commands.

These commands are in viewer mode.

Command
Purpose

show diameter gx virtual-gi all

Displays all the VLAN ID to virtual Gi ID mappings.

show diameter gx virtual-gi vlan-id vlan-id

Displays a particular VLAN ID and virtual ID mapping.

show interface linecard 0 VLAN

Displays the VLAN configuration mode.

show interface linecard 0 subscriber virtual-gi-mode

Displays the status of the virtual Gi mode.


TACACS+ Authentication, Authorization, and Accounting

Information about TACACS+ Authentication, Authorization, and Accounting

Configuring the SCE Platform TACACS+ Client

Managing the User Database

Configuring AAA Login Authentication

Configuring AAA Privilege Level Authorization Methods

Configuring AAA Command Level Authorization Methods

Configuring AAA Accounting

Monitoring TACACS+

Monitoring TACACS+ Users

Information about TACACS+ Authentication, Authorization, and Accounting

Login Authentication

Accounting

Privilege Level Authorization

General AAA Fallback and Recovery Mechanism

About Configuring TACACS+

TACACS+ is a security application that provides centralized authentication of users attempting to gain access to a network element. The implementation of TACACS+ protocol allows customers to configure one or more authentication servers for the SCE platform, providing a secure means of managing the SCE platform, as the authentication server will authenticate each user. This then centralizes the authentication database, making it easier for the customers to manage the SCE platform.

TACACS+ services are maintained in a database on a TACACS+ server running, typically, on a UNIX or Windows NT workstation. You must have access to and must configure a TACACS+ server before the configured TACACS+ features on your network element are available.

The TACACS+ protocol provides authentication between the network element and the TACACS+ ACS, and it can also ensure confidentiality, if a key is configured, by encrypting all protocol exchanges between a network element and a TACACS+ server.

The TACACS+ protocol provides the following three features:

Login authentication

Privilege level authorization

Accounting

Login Authentication

The SCE platform uses the TACACS+ ASCII authentication message for CLI, Telnet and SSH access.

TACACS+ allows an arbitrary conversation to be held between the server and the user until the server receives enough information to authenticate the user. This is usually done by prompting for a username and password combination.

The login and password prompts may be provided by the TACACS+ server, or if the TACACS+ server does not provide the prompts, then the local prompts will be used.

The user log in information (user name and password) is transmitted to the TACACS+ server for authentication. If the TACACS+ server indicates that the user is not authenticated, the user will be re-prompted for the user name and password. The user is re-prompted a user-configurable number of times, after which the failed login attempt is recorded in the SCE platform user log and the telnet session is terminated (unless the user is connected to the console port.)

The SCE platform will eventually receive one of the following responses from the TACACS+ server:

ACCEPT - The user is authenticated and service may begin.

REJECT - The user has failed to authenticate. The user may be denied further access, or will be prompted to retry the login sequence depending on the TACACS+ server.

ERROR - An error occurred at some time during authentication. This can be either at the server or in the network connection between the server and the SCE platform. If an ERROR response is received, the SCE platform will try to use an alternative method\server for authenticating the user.

CONTINUE - The user is prompted for additional authentication information.

If the server is unavailable, the next authentication method is attempted, as explained in General AAA Fallback and Recovery Mechanism.

Accounting

The TACACS+ accounting supports the following functionality:

Each executed command (the command must be a valid one) will be logged using the TACACS+ accounting mechanism (including login and exit commands).

The command is logged both before and after it is successfully executed.

Each accounting message contains the following:

User name

Current time

Action performed

Command privilege level

TACACS+ accounting is in addition to normal local accounting using the SCE platform dbg log.

Privilege Level Authorization

After a successful login the user is granted a default privilege level of 0, giving the user the ability to execute a limited number of commands. Changing privilege level is done by executing the "enable" command. This command initiates the privilege level authorization mechanism.

Privilege level authorization in the SCE platform is accomplished by the use of an "enable" command authentication request. When a user requests an authorization for a specified privilege level, by using the "enable" command, the SCE platform sends an authentication request to the TACACS+ server specifying the requested privilege level. The SCE platform grants the requested privilege level only after the TACACS+ server does the following:

Authenticates the " enable " command password

Verifies that the user has sufficient privileges to enter the requested privilege level.

Once the user privilege level has been determined, the user is granted access to a specified set of commands according to the level granted.

As with login authentication, if the server is unavailable, the next authentication method is attempted, as explained in General AAA Fallback and Recovery Mechanism.

Command Level Authorization

When command level authorization is enabled, each CLI command that is issued must be authorized by the external TACACS server before the system actually executes the command. You can configure the authorization level at which command level authorization is required. For example, you can require command level authorization only at root level.

As with login and privilege level authentication, if the TACACS+ server is unavailable, the regular fall back mechanism will be used.

General AAA Fallback and Recovery Mechanism

The SCE platform uses a fall-back mechanism to maintain service availability in case of an error.

The AAA methods available are:

TACACS+ - AAA is performed by the use of a TACACS+ server, allows authentication, authorization and accounting.

Local - AAA is performed by the use of a local database, allows authentication and authorization.

Enable - AAA is performed by the use of user configured passwords, allows authentication and authorization.

None - no authentication\authorization\accounting is performed.

In the current implementation the order of the methods used isn't configurable but the customer can choose which of the methods are used. The current order is:

TACACS+

Local

Enable

None


Caution If the server goes to AAA fault, the SCE platform will not be accessible until one of the AAA methods is restored. In order to prevent this, it is advisable to use the "none" method as the last AAA method. If the SCE platform becomes inaccessible, the shell function "AAA_MethodsReset" will allow you to delete the current AAA method settings and set the AAA method used to "enable".
To run the "AAA_MethodsReset" shell function, complete the following steps:
1. Connect to AUX with username "root"
2. Run the debug shell: scos_xinetd --service debug-shell --on
3. Use Telnet to access the shell: telnet localhost 2301
4. Run the shell function: AAA_MethodsReset

About Configuring TACACS+

The following is a summary of the procedure for configuring TACACS+. All steps are explained in detail in the remainder of this section.

1. Configure the remote TACACS+ servers.

Configure the remote servers for the protocols. Keep in mind the following guidelines

Configure the encryption key that the server and client will use.

The maximal user privilege level and enable password (password used when executing the enable command) should be provided.

The configuration should always include the root user, giving it the privilege level of 15.

Viewer (privilege level 5) and superuser (privilege level 10) user IDs should be established at this time also.

1. For complete details on server configuration, refer to the appropriate configuration guide for the particular TACACS+ server that you will be using.

2. Configure the SCE client to work with TACACS+ server:

hostname of the server

port number

shared encryption key (the configured encryption key must match the encryption key configured on the server in order for the client and server to communicate.)

3. (Optional) Configure the local database, if used.

add new users

If the local database and TACACS+ are both configured, it is recommended to configure the same user names in both TACACS+ and the local database. This will allow the users to access the SCE platform in case of TACACS+ server failure.


Note If TACACS+ is used as the login method, the TACACS+ username is used automatically in the enable command. Therefore, it is important to configure the same usernames in both TACACS+ and the local database so that the enable command can recognize this username.


specify the password

define the privilege level

4. Configure the authentication methods on the SCE platform.

login authentication methods

privilege level authorization methods

command level authorization methods

5. Review the configuration.

Use the " show running-config " command to view the configuration.

Configuring the SCE Platform TACACS+ Client

Adding a New TACACS+ Server Host

Removing a TACACS+ Server Host

Configuring the Global Default Key

Configuring the Global Default Timeout

The user must configure the remote servers for the TACACS+ protocol. Then the SCE platform TACACS+ client must be configured to work with the TACACS+ servers. The following information must be configured:

TACACS+ server hosts definition — a maximum of three servers is supported.

For each sever host, the following information can be configured:

hostname (required)

port

encryption key

timeout interval

Default encryption key (optional) — A global default encryption key may be defined. This key is defined as the key for any server host for which a key is not explicitly configured when the server host is defined.

If the default encryption key is not configured, a default of no key is assigned to any server for which a key is not explicitly configured.

Default timeout interval (optional) — A global default timeout interval may be defined. This timeout interval is defined as the timeout interval for any server host for which a timeout interval is not explicitly configured when the server host is defined.

If the default timeout interval is not configured, a default of five seconds is assigned to any server for which a timeout interval is not explicitly configured.

The procedures for configuring the SCE platform TACACS+ client are explained in the following sections:

Adding a New TACACS+ Server Host

Removing a TACACS+ Server Host

Configuring the Global Default Key

Configuring the Global Default Timeout

Adding a New TACACS+ Server Host

Use this command to define a new TACACS+ server host that is available to the SCE platform TACACS+ client.

The Service Control solution supports a maximum of three TACACS+ server hosts.

Options

The following options are available:

host-name — name of the server

port number — TACACS+ port number

Default = 49

timeout interval — time in seconds that the server waits for a reply from the server host before timing out

Default = 5 seconds or user-configured global default timeout interval (see To define the global default timeout, do the following:.)

key-string — encryption key that the server and client will use when communicating with each other. Make sure that the specified key is actually configured on the TACACS+ server host.

Default = no key or user-configured global default key (see To define a global default key, do the following:.)

From the SCE(config)# prompt, type:

Command
Purpose

tacacs-server host host-name [port portnumber] [timeout timeout-interval] [key key-string]

Define a new TACACS+ server host that is available to the SCE platform TACACS+ client.


Removing a TACACS+ Server Host

Options

The following options are available:

host-name — name of the server to be deleted

From the SCE(config)# prompt, type:

Command
Purpose

no tacacs-server host host-name

Removes a TACACS+ server host.


Configuring the Global Default Key

Use this command to define the global default key for the TACACS+ server hosts. This default key can be overridden for a specific TACACS+ server host by explicitly configuring a different key for that TACACS+ server host.

Options

The following options are available:

key-string — default encryption key that all TACACS+ servers and clients will use when communicating with each other. Make sure that the specified key is actually configured on the TACACS+ server hosts.

Default = no encryption

To define a global default key, do the following:

From the SCE(config)# prompt, type:

Command
Purpose

tacacs-server key key-string

Defines the global default key for the TACACS+ server hosts.


To clear a global default key, do the following:

From the SCE(config)# prompt, type:

Command
Purpose

no tacacs-server key

No global default key is defined. Each TACACS+ server host may still have a specific key defined. However, any server host that does not have a key explicitly defined (uses the global default key) is now configured to use no key.


Configuring the Global Default Timeout

Use this command to define the global default timeout interval for the TACACS+ server hosts. This default timeout interval can be overridden for a specific TACACS+ server host by explicitly configuring a different timeout interval for that TACACS+ server host.

Options

The following options are available:

timeout interval — default time in seconds that the server waits for a reply from the server host before timing out.

Default = 5 seconds

To define the global default timeout, do the following:

From the SCE(config)# prompt, type:

Command
Purpose

tacacs-server timeout timeout-interval

Defines global default timeout.


To clear the global default timeout, do the following:

From the SCE(config)# prompt, type:

Command
Purpose

no tacacs-server timeout

No global default timeout interval is defined. Each TACACS+ server host may still have a specific timeout interval defined. However, any server host that does not have a timeout interval explicitly defined (uses the global default timeout interval) is now configured to a five second timeout interval.


Managing the User Database

TACACS+ maintains a local user database. Up to 100 users can be configured in this local database, which includes the following information for all users:

Username

Password — may configured as encrypted or unencrypted

Privilege level

The procedures for managing the local user database are explained in the following sections:

Adding a New User to the Local Database

Defining the User Privilege Level

Adding a New User with Privilege Level and Password

Deleting a User

Adding a New User to the Local Database

Use these commands to add a new user to the local database. Up to 100 users may be defined.

How to Add a User with a Clear Text Password

How to Add a User with No Password

How to Add a User with an MD5 Encrypted Password Entered in Clear Text

How to Add a User with an MD5 Encrypted Password Entered as an MD5 Encrypted String

Options

The password is defined with the username. There are several password options:

No password — Use the nopassword keyword.

Password — Password is saved in clear text format in the local list.

Use the password parameter.

Encrypted password — Password is saved in encrypted (MD5) form in the local list. Use the secret keyword.

Password may be defined by either of the following methods:

Specify a clear text password, which is saved in MD5 encrypted form

Specify an MD5 encryption string, which is saved as the user MD5-encrypted secret password

The following options are available:

name — name of the user to be added

password — a clear text password. May be saved in the local list in either of two formats:

as clear text

in MD5 encrypted form if the secret keyword is used

encrypted-secret — an MD5 encryption string password

The following keywords are available:

nopassword — There is no password associated with this user

secret — the password is saved in MD5 encrypted form. Use with either of the following keywords to indicate the format of the password as entered in the command:

0 — use with the password option to specify a clear text password that will be saved in MD5 encrypted form

5 — use with the encrypted-secret option to specify an MD5 encryption string that will be saved as the user MD5-encrypted secret password

How to Add a User with a Clear Text Password

From the SCE(config)# prompt, type:

Command
Purpose

username name password password

Adds a user with a clear text password.


How to Add a User with No Password

From the SCE(config)# prompt, type:

Command
Purpose

username name nopassword

Adds a user with no password.


How to Add a User with an MD5 Encrypted Password Entered in Clear Text

From the SCE(config)# prompt, type:

Command
Purpose

username name secret 0 password

Adds a user with an MD5 encrypted password entered in clear text.


How to Add a User with an MD5 Encrypted Password Entered as an MD5 Encrypted String

From the SCE(config)# prompt, type:

Command
Purpose

username name secret 5 encrypted-secret

Adds a user with an MD5 encrypted password entered as an MD5 encrypted string.


Defining the User Privilege Level

Privilege level authorization in the SCE platform is accomplished by the use of an " enable " command authentication request. When a user requests an authorization for a specified privilege level, by using the " enable " command, the SCE platform sends an authentication request to the TACACS+ server specifying the requested privilege level. The SCE platform grants the requested privilege level only after the TACACS+ server authenticates the " enable " command password and verifies that the user has sufficient privileges the enter the requested privilege level.

Options

The following options are available:

name — name of the user whose privilege level is set

level — the privilege level permitted to the specified user. These levels correspond to the CLI authorization levels, which are entered via the enable command:

0 — User

10 — Admin

15 (default) — Root

From the SCE(config)# prompt, type:

Command
Purpose

username name privilege level

Defines user privilege level.


Adding a New User with Privilege Level and Password

Use these commands to define a new user, including password and privilege level, in a single command.


Note In the config files (running config and startup config ), this command will appear as two separate commands.


How to Add a User with a Privilege Level and a Clear Text Password

How to Add a User with a Privilege Level and an MD5 Encrypted Password Entered in Clear Text

How to Add a User with a Privilege Level and an MD5 Encrypted Password Entered as an MD5 Encrypted String

Options

The following options are available:

name — name of the user whose privilege level is set

level — the privilege level permitted to the specified user. These levels correspond to the CLI authorization levels, which are entered via the enable command:

0 — User

10 — Admin

15 (default) — Root

password — a clear text password. May be saved in the local list in either of two formats:

as clear text I

n MD5 encrypted form if the secret keyword is used

encrypted-secret — an MD5 encryption string password

The following keywords are available:

secret — the password is saved in MD5 encrypted form. Use with either of the following keywords to indicate the format of the password as entered in the command:

0 — use with the password option to specify a clear text password that will be saved in MD5 encrypted form

5 = use with the encrypted-secret option to specify an MD5 encryption string that will be saved as the user MD5-encrypted secret password

How to Add a User with a Privilege Level and a Clear Text Password

From the SCE(config)# prompt, type:

Command
Purpose

username name privilege level password password

Adds a user with a privilege level and a clear text password.


How to Add a User with a Privilege Level and an MD5 Encrypted Password Entered in Clear Text

From the SCE(config)# prompt, type:

Command
Purpose

username name privilege level secret 0 password

Adds a user with a privilege level and an MD5 encrypted password entered in clear text.


How to Add a User with a Privilege Level and an MD5 Encrypted Password Entered as an MD5 Encrypted String

From the SCE(config)# prompt, type:

Command
Purpose

username name privilege level secret 5 encrypted-secret

Adds a user with a privilege level and an MD5 encrypted password entered as an MD5 encrypted string.


Deleting a User

Options

The following options are available:

name — name of the user to be deleted

From the SCE(config)# prompt, type:

Command
Purpose

no username name

Deletes a user.


Configuring AAA Login Authentication

There are two features to be configured for login authentication:

Maximum number of permitted Telnet login attempts

The authentication methods used at login (see General AAA Fallback and Recovery Mechanism.)

The procedures for configuring login authentication are explained in the following sections:

Configuring Maximum Login Attempts

Configuring the AAA Login Authentication Methods

Configuring Maximum Login Attempts

Use this command to set the maximum number of login attempts that will be permitted before the session is terminated.

Options

The following options are available:

number-of-attempts — The maximum number of login attempts that will be permitted before the telnet session is terminated.

This is relevant only for Telnet sessions. From the local console, the number of re-tries is unlimited.

Default = three

From the SCE(config)# prompt, type:

Command
Purpose

aaa authentication attempts login number-of-attempts

Configures maximum login attempts.


Configuring the AAA Login Authentication Methods

You can configure "backup" login authentication methods to be used if failure of the primary login authentication method (see General AAA Fallback and Recovery Mechanism).

Use this command to specify which login authentication methods are to be used, and in what order of preference.

How to Specify the Login Authentication Methods

How to Delete the Login Authentication Methods List

Options

The following options are available:

method — the login authentication methods to be used. You may specify up to four different methods, in the order in which they are to be used.

group TACACS+ — Use TACACS+ authentication.

local — Use the local username database for authentication

enable (default) — Use the " enable " password for authentication

none — Use no authentication.

How to Specify the Login Authentication Methods

From the SCE(config)# prompt, type:

Command
Purpose

aaa authentication login default method1 [method2...]

Specifies login authentication methods.

You may list a maximum of four methods; all four methods explained above. List them in the order of priority.


How to Delete the Login Authentication Methods List

From the SCE(config)# prompt, type:

Command
Purpose

no aaa authentication login default

Deletes login authentication methods list.

If the login authentication methods list is deleted, the default login authentication method only (enable password) will be used. TACACS+ authentication will not be used.


Configuring AAA Privilege Level Authorization Methods

How to Specify AAA Privilege Level Authorization Methods

How to Delete the AAA Privilege Level Authorization Methods List

Options

The following options are available:

method — the login authorization methods to be used. You may specify up to four different methods, in the order in which they are to be used.

group TACACS+ — Use TACACS+ authorization.

local — Use the local username database for authorization

enable (default) — Use the " enable " password for authorization

none — Use no authorization.

How to Specify AAA Privilege Level Authorization Methods

From the SCE(config)# prompt, type:

Command
Purpose

aaa authorization enable default method1 [method2...]

Specifies AAA privilege level authorization methods.

You may list a maximum of four methods; all four methods explained above. List them in the order of priority.


How to Delete the AAA Privilege Level Authorization Methods List

From the SCE(config)# prompt, type:

Command
Purpose

no aaa authorization enable default

Deletes AAA privilege level authorization methods list.

If the privilege level authorization methods list is deleted, the default login authentication method only (enable password) will be used. TACACS+ authentication will not be used.


Configuring AAA Command Level Authorization Methods

How to Specify AAA Command Level Authorization Methods

How to Delete the AAA Command Level Authorization Methods List

Options

The following options are available:

level — The privilege level for which to enable the TACACS+ command level authorizarion (0, 5, 10, 15)

method — the command authorization methods to be used. You may specify up to two methods, in the order in which they are to be used.

group TACACS+ — Use TACACS+ authorization.

none — Use no authorization.

How to Specify AAA Command Level Authorization Methods

From the SCE(config)# prompt, type:

Command
Purpose

aaa authorization command level default method1 [method2]]

Specifies AAA command level authorization methods.

You may list a maximum of two methods. List them in the order of priority.


How to Delete the AAA Command Level Authorization Methods List

From the SCE(config)# prompt, type:

Command
Purpose

no aaa authorization command level default

Deletes AAA command level authorization methods list.

If the command level authorization methods list is deleted, the default login authentication method only (enable password) will be used. TACACS+ authentication will not be used.


Configuring AAA Accounting

Use this command to enable or disable TACACS+ accounting.

How to Enable AAA Accounting

How to Disable AAA Accounting

If TACACS+ accounting is enabled, the SCE platform sends an accounting message to the TACACS+ server after every command execution. The accounting message is logged in the TACACS+ server for the use of the network administrator.

By default, TACACS+ accounting is disabled.

Options

The following options are available:

level — The privilege level for which to enable the TACACS+ accounting

How to Enable AAA Accounting

From the SCE(config)# prompt, type:

Command
Purpose

aaa authentication accounting commands level default stop-start group tacacs+

Enables AAA accounting.

The start-stop keyword (required) indicates that the accounting message is sent at the beginning and the end (if the command was successfully executed) of the execution of a CLI command.


How to Disable AAA Accounting

From the SCE(config)# prompt, type:

Command
Purpose

aaa authentication accounting commands level default

Disables AAA accounting.


Monitoring TACACS+

Displaying Statistics for TACACS+ Servers

Displaying Statistics, Keys and Timeouts for TACACS+ Servers

Monitoring TACACS+ Users

Displaying Statistics for TACACS+ Servers

From the SCE# prompt, type:

Command
Purpose

show tacacs

Displays statistics for TACACS+ servers.


Displaying Statistics, Keys and Timeouts for TACACS+ Servers

From the SCE# prompt, type:

Command
Purpose

show tacacs all

Displays statistics, keys, and timeouts for TACACS+ servers.

Note that, although most show commands are accessible to viewer level users, the ' all ' option is available only at the admin level. Use the command ' enable 10 ' to access the admin level.


Monitoring TACACS+ Users

Use this command to display the users in the local database, including passwords.

From the SCE# prompt, type:

Command
Purpose

show users

Displays the users in the local database, including passwords.

Note that, although most show commands are accessible to viewer level users, this command is available only at the admin level. Use the command ' enable 10 ' to access the admin level.


Configuring Access Control Lists (ACLs)

Adding Entries to an ACL

Removing an ACL

Defining a Global ACL

The SCE platform can be configured with Access Control Lists (ACLs), which are used to permit or deny incoming connections on the management interface. An access list is an ordered list of entries, each consisting of an IP address and an optional wildcard "mask" defining an IP address range, and a permit/deny field.

The order of the entries in the list is important. The default action of the first entry that matches the connection is used. If no entry in the Access List matches the connection, or if the Access List is empty, the default action is deny.

Configuration of system access is done in two stages:

1. Creating an access list. (Adding Entries to an ACL).

2. Associating the ACL with a management service. (See Defining a Global ACL.)

Creating an access list is done entry by entry, from the first to the last.

When the system checks for an IP address on an access list, the system checks each line in the ACL for the IP address, starting at the first entry and moving towards the last entry. The first match that is detected (that is, the IP address being checked is found within the IP address range defined by the entry) determines the result, according to the permit/deny flag in the matched entry. If no matching entry is found in the ACL, access is denied.

You can create up to 99 ACLs. ACLs can be associated with system access on the following levels:

Global (IP) level: If a global list is defined using the ip access-class command, when a request comes in, the SCE platform first checks if there is permission for access from that IP address. If not, the SCE does not respond to the request. Configuring the SCE platform to deny a certain IP address would preclude the option of communicating with that address using any IP-based protocol including Telnet, FTP, ICMP, RPC, SSH, and SNMP. The basic IP interface is low-level, blocking the IP packets before they reach the interfaces.

Service level: Access to each management service (Telnet, SNMP, and SSH) can be restricted to an ACL. Interface-level lists are, by definition, a subset of the global list defined. If access is denied at the global level, the IP will not be allowed to access using one of the interfaces. Once an ACL is associated with a specific management service, that service checks the ACL to find out if there is permission for a specific external IP address trying to access the management interface.

Use the following CLI commands to assign an ACL to the specified management service:

Telnet-access-class in

SSH—ip ssh access-class

SNMP—snmp-server community

It is possible to configure several management services to the same ACL, if this is the desired behavior of the SCE platform.

If no ACL is associated to a management service or to the global IP level, access is permitted from all IP addresses.


Note The SCE Platform will respond to ping commands only from IP addresses that are allowed access. Pings from a non-authorized address will not receive a response from the SCE platform, as ping uses ICMP protocol.


Options

The following options are available:

number — the ID number assigned to the Access Control List

ip-address — the IP address of the interface to be permitted or denied. Enter in x.x.x.x format.

ip-address/mask — configures a range of addresses in the format x.x.x.x y.y.y.y where x.x.x.x specifies the prefix bits common to all IP addresses in the range, and y.y.y.y is a wildcard-bits mask specifying the bits that are ignored. In this notation, `0' means bits to ignore.

The following keywords are available:

permit — the specified IP addresses have permission to access the SCE platform.

deny — the specified IP addresses are denied access to the SCE platform.

Adding Entries to an ACL


Step 1 Type configure and press Enter.

Enables Global Configuration mode.

Step 2 Enter the desired IP address or addresses.

To configure one IP address type:

access-list number permit|deny ip-address and press Enter.

To configure more than one IP address type:

access-list number permit|deny ip-address/mask and press Enter.

When you add a new entry to an ACL, it is always added to the end of the list.


Adding Entries to an ACL: Example

The following example adds an entry to the access list number 1, that permits access only to IP addresses in the range of 10.1.1.0-10.1.1.255.

SCE(config)#access-list 1 permit 10.1.1.0 0.0.0.255

Removing an ACL

Use this command to remove an ACL with all its entries.

From the SCE(config)# prompt, type:

Command
Purpose

no access-list number

Removes the specified ACL with all its entries.


Defining a Global ACL

A global ACL for permits or denies all traffic to the SCE platform.

From the SCE(config)# prompt, type:

Command
Purpose

ip access-class number

Applies the specified ACL to all traffic attempting to access the SCE platform, rather than to a specific type of traffic, such as Telnet traffic.


Managing the Telnet Interface

Preventing Telnet Access

Assigning an ACL to the Telnet Interface

Configuring the Telnet Timeout

This section discusses the Telnet interface of the SCE platform. A Telnet session is the most common way to connect to the SCE platform CLI interface.

You can set the following parameters for the Telnet interface:

Enable/disable the interface

Assign an ACL to permit or deny incoming connections.

Timeout for Telnet sessions, that is, if there is no activity on the session, how long the SCE platform waits before automatically cutting off the Telnet connection.

The following commands are relevant to Telnet interface:

access-class in

line vty

[no] access list

[no] service telnetd

[no] timeout

show line vty access-class in

show line vty timeout

Preventing Telnet Access

Use this command to disable access by Telnet altogether.

From the SCE(config)# prompt, type:

Command
Purpose

no service telnetd

Disables access by Telnet.

Current Telnet sessions are not disconnected, but no new Telnet sessions are allowed.


Assigning an ACL to the Telnet Interface

From the SCE(config-line)# prompt, type:

Command
Purpose

access-class acl-number in

The specified ACL controls access to the Telnet interface.

acl-number is the ID number of an existing ACL.


Assigning an ACL to the Telnet Interface: Example

The following example shows how to assign ACL #1 to the Telnet interface.

SCE#configure 
SCE(config)#line vty 0 
SCE(config-line)#access-class 1 in

Removing the ACL Assignment from the Telnet Interface

From the SCE(config-line)# prompt, type:

Command
Purpose

no access-class in

Removes the ACL assignment from the Telnet interface, so that any IP address may now access the Telnet interface.


Configuring the Telnet Timeout

The SCE platform supports timeout of inactive Telnet sessions.

Options

The following options are available:

timeout — The length of time in minutes before an inactive Telnet session will be timed-out.

Default — 30 minutes

From the SCE(config-line)# prompt, type:

Command
Purpose

timeout timeout

Configures Telnet timeout.


Configuring the SSH Server

The SSH Server

Key Management

Managing the SSH Server

Monitoring the Status of the SSH Server

The SSH Server

A shortcoming of the standard telnet protocol is that it transfers password and data over the net unencrypted, thus compromising security. Where security is a concern, using a Secure Shell (SSH) server rather than telnet is recommended.

An SSH server is similar to a telnet server, but it uses cryptographic techniques that allow it to communicate with any SSH client over an insecure network in a manner which ensures the privacy of the communication. CLI commands are executed over SSH in exactly the same manner as over telnet.

The SSH server supports both the SSHv1 and SSHv2 protocols. You can disable SSHv1, so that only SSHv2 is running.

The SSH server supports the following encryption ciphers:

aes256-ctr, aes192-ctr, aes128-ctr (RFC-4344, section 4).

3des-cbc, blowfish-cbc, aes256-cbc, aes192-cbc, aes128-cbc, arcfour, cast128-cbc (RFC-4253, section 6.3)

arcfour128, arcfour256 (RFC-4345, section 4).

rijndael-cbc@lysator.liu.se (as provided by OpenSSH 4.7p1).

An ACL can be configured for SSH as for any other management protocol, limiting SSH access to a specific set of IP addresses (see Configuring Access Control Lists (ACLs))

Key Management

Each SSH server should define a set of keys (DSA2, RSA2 and RSA1) to be used when communicating with various clients. The key sets are pairs of public and private keys. The server publishes the public key while keeping the private key in non-volatile memory, never transmitting it to SSH clients. Note that the keys are kept on the tffs0 file system, which means that a person with knowledge of the `enable' password can access both the private and public keys. The SSH server implementation provides protection against eavesdroppers who can monitor the management communication channels of the SCE platform, but it does not provide protection against a user with knowledge of the `enable' password.

Key management is performed by the user via a special CLI command. A set of keys must be generated at least once before enabling the SSH server.

Size of the encryption key is always 2048 bits.

Managing the SSH Server

Use these commands to manage the SSH server. These commands do the following:

Generate an SSH key set

Enable/disable the SSH server

Enable/disable SSHv1. (Disabling SSHv1 allows you to run SSHv2 only.)

Delete existing SSH keys

Generating a Set of SSH Keys

Remember that you must generate a set of SSH keys before you enable the SSH server.

From the SCE(config)# prompt, type:

Command
Purpose

ip ssh key generate

Generates a new SSH key set and immediately saves it to non-volatile memory. (Key set is not part of the configuration file). Key size is always 2048 bits.


Enabling the SSH Server

SSH allows you to login only when the user password and AAA authentication

are configured.

Configure at least one user name and password.

SCE8000(config)# username <username> password <password>

Configure AAA authentication for login.

SCE8000(config)# aaa authentication login default none

From the SCE(config)# prompt, type:

Command
Purpose

ip ssh

Enables SSH server.


Disabling the SSH Server

From the SCE(config)# prompt, type:

Command
Purpose

no ip ssh

Disables SSH server.


Running Only SSHv2


Step 1 From the SCE(config)# prompt, type ip ssh and press Enter.

Step 2 From the SCE(config)# prompt, type no ip ssh sshv1 and press Enter

To re-enable SSHv1, use the command ip ssh SSHv1.


Assigning an ACL to the SSH Server

From the SCE(config)# prompt, type:

Command
Purpose

ip ssh access-class acl-number

The specified ACL controls access to the SSH server.

acl-number is the ID number of an existing ACL


Removing the ACL Assignment from the SSH Server

rom the SCE(config)# prompt, type:

Command
Purpose

no ip ssh access-class

Removes the ACL assignment from the SSH server, so that any IP address may now access the SSH server.


Deleting the Existing SSH Keys

From the SCE(config)# prompt, type:

Command
Purpose

ip ssh key remove

Removes the existing SSH key set from non-volatile memory.


If the SSH server is currently enabled, it will continue to run, since it only reads the keys from non-volatile memory when it is started. However, if the startup-configuration specifies that the SSH server is enabled, the SCE platform will not be able to start the SSH server on startup if the keys have been deleted. To avoid this situation, after executing this command, always do one of the following before the SCE platform is restarted (using reload ):

Generate a new set of keys.

Disable the SSH server and save the configuration.

Monitoring the Status of the SSH Server

Use this command to monitor the status of the SSH sever, including current SSH sessions.

From the SCE> prompt, type:

Command
Purpose

show ip ssh

Monitors the status of SSH server.


Configuring and Managing the SNMP Interface

About the SNMP Interface

Enabling the SNMP Interface

Configuring SNMP Community Strings

Configuring SNMP Notifications

About the SNMP Interface

This section explains how to configure the SNMP agent parameters. It also provides a brief overview of SNMP notifications and relevant CLI commands.

SNMP Protocol

Security Considerations

About CLI

About MIBs

Configuration via SNMP

SNMP Protocol

SNMP (Simple Network Management Protocol) is a set of protocols for managing complex networks. SNMP works by sending messages, called protocol data units (PDUs), to different parts of a network. SNMP-compliant devices, called agents, store data about themselves in Management Information Bases (MIBs) and return this data to the SNMP requesters.

SCE platform supports the original SNMP protocol (also known as SNMPv1), and a newer version called Community-based SNMPv2 (also known as SNMPv2C).

SNMPv1 — is the first version of the Simple Network Management Protocol, as defined in RFCs 1155 and 1157, and is a full Internet standard. SNMPv1 uses a community-based form of security.

SNMPv2c — is the revised protocol, which includes improvements to SNMPv1 in the areas of protocol packet types, transport mappings, and MIB structure elements but using the existing SNMPv1 administration structure. It is defined in RFC 1901, RFC 1905, and RFC 1906.

SCE platform implementation of SNMP supports all MIB II variables, as described in RFC 1213, and defines the SNMP traps using the guidelines described in RFC 1215.

The SNMPv1 and SNMPv2C specifications define the following basic operations that are supported by SCE platform. Table 5-3 lists the request types and descriptions.

Table 5-3 Request Types 

Request Type
Description
Remarks

Set Request

Writes new data to one or more of the objects managed by an agent.

Set operations immediately affect the SCE platform running-config but do not affect the startup config

Get Request

Requests the value of one or more of the objects managed by an agent.

 

Get Next Request

Requests the Object Identifier(s) and value(s) of the next object(s) managed by an agent.

 

Get Response

Contains the data returned by an agent.

 

Trap

Sends an unsolicited notification from an agent to a manager, indicating that an event or error has occurred on the agent system

SCE platform may be configured to send either SNMPv1 or SNMPv2 style traps.

Get Bulk Request

Retrieves large amounts of object information in a single Request / response transaction. GetBulk behaves as if many iterations of GetNext request/responses were issued, except that they are all performed in a single request/response.

This is newly defined SNMPv2c message.


Security Considerations

By default, the SNMP agent is disabled for both read and write operations. When enabled, SNMP is supported over the management port only (in-band management is not supported).

In addition, the SCE platform supports the option to configure community of managers for read-write accessibility or for read-only accessibility. Furthermore, an ACL may be associated with the SNMP agent by assigning it to one of the community strings to allow SNMP management to a restricted set of manager IP addresses. If different ACLs are assigned to different community strings, access by all community strings is controlled by all assigned ACLs. Assigning different ACLs to different community strings is not supported

About CLI

CLI Commands for Configuring SNMP

CLI Commands for Monitoring SNMP

The SCE platform supports the CLI commands that control the operation of the SNMP agent. All the SNMP commands are available in Admin authorization level. The SNMP agent is disabled by default and any SNMP configuration command enables the SNMP agent (except where there is an explicit disable command).

CLI Commands for Configuring SNMP

Following is a list of CLI commands available for configuring SNMP. These are Global Configuration mode commands.

snmp-server enable

no snmp-server

[no] snmp-server community [all]

[no | default] snmp-server enable traps

[no] snmp-server host [all]

[no] snmp-server contact

[no] snmp-server location

CLI Commands for Monitoring SNMP

Following is a list of CLI commands available for monitoring SNMP. These are Viewer mode commands, and are available when the SNMP agent is enabled:

show snmp (available when SNMP agent is disabled)

show snmp community

show snmp contact

show snmp enabled

show snmp host

show snmp location

show snmp MIB (available when SNMP agent enabled and community was set)

show snmp traps

About MIBs

MIBs (Management Information Bases) are databases of objects that can be monitored by a network management system (NMS). SNMP uses standardized MIB formats that allow any SNMP tools to monitor any device defined by a MIB.

For further information concerning MIBs used by the Cisco SCE8000 platform, see the Cisco Service Control MIBs

Configuration via SNMP

SCE platform supports a limited set of variables that may be configured via SNMP (read-write variables). Setting a variable via SNMP (as via the CLI) takes effect immediately and affects only the running-configuration. To make this configuration stored for next reboots (startup-configuration) the user must specify it explicitly via CLI or via SNMP using the Cisco enterprise MIB objects.

It should be noted also that the SCE platform takes the approach of a single configuration database with multiple interfaces that may change this database. Therefore, executing the copy running-config startup-config command via CLI or SNMP makes permanent all the changes made by either SNMP or CLI.

Enabling the SNMP Interface

Use this command to explicitly enable the SNMP interface.

The SNMP interface is implicitly enabled when any snmp-server command is executed to configure any SNMP parameter. For more information on configuring and managing the SNMP parameters, including hosts, communities, contact, location, and trap destination hosts, see Configuring and Managing the SNMP Interface.

How to Enable the SNMP Interface

How to Disable the SNMP Interface

How to Enable the SNMP Interface

You must define at least one community string to allow SNMP access. For complete information on community strings see Configuring SNMP Community Strings.

From the SCE(config)# prompt, type:

Command
Purpose

snmp-server enable

Enables SNMP interface.


How to Disable the SNMP Interface

From the SCE(config)# prompt, type:

Command
Purpose

no snmp-server

Disables SNMP interface.


Configuring SNMP Community Strings

Defining a Community String

Removing a Community String

Displaying the Configured Community Strings

To enable SNMP management, you must configure SNMP community strings to define the relationship between the SNMP manager and the agent.

After receiving an SNMP request, the SNMP agent compares the community string in the request to the community strings that are configured for the agent. The requests are valid under the following circumstances:

SNMP Get , Get-next , and Get-bulk requests are valid if the community string in the request matches the read-only community.

SNMP Get , Get-next , Get-bulk and Set requests are valid if the community string in the request matches the agent's read-write community.

Defining a Community String

Options

The following options are available:

community-string — a security string that identifies a community of managers who are permitted to access the SNMP server

acl-number—(Optional) ID number of the ACL to be assigned to the SNMP interface. It should list the IP addresses of all SNMP managers permitted to access the SNMP server.


Note Assigning different ACLs to different community strings is not supported. If you specify an ACL in this command, it is assigned to the SNMP server globally, not just to the specified community string. For example, if you configure two community strings and assign a different ACL to each, access to the SNMP agent for both communities is controlled by both ACLs.


If no ACL is specified, all IP addresses can access the SNMP service. For more information about ACLs, see Configuring Access Control Lists (ACLs).

The following keywords are available:

ro — read only (default accessibility)

rw — read and write

From the SCE(config)# prompt, type:

Command
Purpose

snmp-server community community-string ro|rw [acl-number]

Defines a community string.

If you specify ACLs for any communities, all assigned ACLs in conjunction will control access for all communities.

Repeat the command as necessary to define all community strings.


Defining a Community String: Example

This example shows how to configure a community string called "mycommunity" with read-only rights. ACL "1" will be assigned to the SNMP server and all configured community strings, not just "mycommunity".

Since read-only is the default, it does not need to be defined explicitly.

SCE(config)#snmp-server community mycommunity 1

Removing a Community String

From the SCE(config)# prompt, type:

Command
Purpose

no snmp-server community community-string

Removes a community string.


Removing a Community String: Example

The following example shows how to remove a community string called "mycommunity". If an ACL was assigned to this community string, it is also removed.

SCE(config)#no snmp-server community mycommunity

Displaying the Configured Community Strings

From the SCE> prompt, type:

Command
Purpose

show snmp-server community community-string

Displays configured community strings


Displaying the Configured Community Strings: Example

The following example shows how to display the configured SNMP communities.

SCE>show snmp community 
Community: public, Access Authorization: RO, Access List Index: 1 
SCE>

Configuring SNMP Notifications

Use these commands to configure:

The destinations that will receive SNMP notifications (hosts)

Which types of notifications will be sent (traps)

Notifications are unsolicited messages that are generated by the SNMP agent that resides inside the SCE platform when an event occurs. When the Network Management System receives the notification message, it can take suitable actions, such as logging the occurrence or ignoring the signal.

By default, the SCE platform is not configured to send any SNMP notifications. You must define the Network Management System to which the SCE platform should send notifications. (See the table below, Configurable Notifications, for a list of configurable notifications). Whenever one of the events that trigger notifications occurs in the SCE platform, an SNMP notification is sent from the SCE platform to the list of IP addresses that you define.

SCE platform supports two general categories of notifications:

Standard SNMP notifications - As defined in RFC1157 and using the conventions defined in RFC1215.

Proprietary service control enterprise notifications - As defined in the service control proprietary MIB (See Table A-20).

After a host or hosts are configured to receive notifications, by default, the SCE platform sends to the host or hosts all the notifications supported by the SCE platform except for the AuthenticationFailure notification. The SCE platform provides the option to enable or disable the sending of this notification, as well as some of the SCE enterprise notifications, explicitly.

SCE platform can be configured to generate either SNMPv1 style or SNMPv2c style notifications. By default, the SCE platforms sends SNMPv1 notifications.

Following are some sample procedures illustrating how to do the following:

Configure hosts (NMS) to which the SNMP agent should send notifications

Remove/disable a host (NMS) from receiving notifications

Enable the SNMP agent to send authentication-failure notifications

Enable the SNMP agent to send enterprise notifications

Reset all notifications to the default setting

Defining SNMP Hosts

Use this command to define the hosts that will receive notifications from the SCE platform.

How to Configure the SCE Platform to Send Notifications to a Host (NMS)

How to Configure the SCE Platform to Stop Sending Notifications to a Host

Configuring SNMP Traps

Options

The following options are available:

ip-address — the IP address of the SNMP server host

community-string — a security string that identifies a community of managers who are permitted to access the SNMP server

version — SNMP version running in the system. Can be set to 1 or 2c.

Default — 1 (SNMPv1)

How to Configure the SCE Platform to Send Notifications to a Host (NMS)

At the SCE(config)# prompt, type:

Command
Purpose

snmp-server host ip-address community-string

Configures SCE platform to send notifications to a host (NMS).

If the version is not specified, SNMPv1 is assumed.

Only one host can be specified per command. To define multiple hosts, execute one command for each host.


Configuring the SCE Platform to Send Notifications to Multiple Hosts: Example

The following example shows how to configure the SCE platform to send SNMPv1 notifications to several hosts.

SCE(config)#snmp-server host 10.10.10.10 mycommunity 
SCE(config)#snmp-server host 20.20.20.20 mycommunity 
SCE(config)#snmp-server host 30.30.30.30 mycommunity 
SCE(config)#snmp-server host 40.40.40.40 mycommunity

How to Configure the SCE Platform to Stop Sending Notifications to a Host

At the SCE(config)# prompt, type:

Command
Purpose

no snmp-server host ip-address

Configures SCE platform to stop sending notifications to a host.


Configuring the SCE Platform to Stop Sending Notifications to a Host: Example

The following example shows how to remove the host with the IP Address: "192.168.0.83".

SCE(config)#no snmp-server host 192.168.0.83

Configuring SNMP Traps

Use this command to configure the notifications that will be sent to the defined host.

How to Enable the SNMP Server to Send Authentication Failure Notifications

How to Enable the SNMP Server to Send All Enterprise Notifications

How to Enable the SNMP Server to Send a specific Enterprise Notification

How to Restore All Notifications to the Default Status

Options

The following options are available:

snmp — optional parameter that specifies that all or specific snmp traps should be enabled or disabled.

By default, snmp traps are disabled.

snmp trap name — optional parameter that specifies a specific snmp trap that should be enabled or disabled.

Currently the only accepted value for this parameter is Authentication .

enterprise — optional parameter that specifies that all or specific enterprise traps should be enabled or disabled.

By default, enterprise traps are enabled.

enterprise trap name — optional parameter that specifies a specific snmp trap that should be enabled or disabled.

Values: attack, chassis, link-bypass, logger, operational-status, port-operational-status, pull-request-failure, RDR-formatter, session, SNTP, subscriber, system-reset, telnet, vas-traffic-forwarding

Use these parameters as follows:

To enable/disable all traps of one type: Specify only snmp or enterprise .

To enable/disable only one specific trap: Specify snmp or enterprise with the additional trap name parameter naming the desired trap.

To enable/disable all traps: Do not specify either snmp or enterprise .

How to Enable the SNMP Server to Send Authentication Failure Notifications

At the SCE(config)# prompt, type:

Command
Purpose

snmp-server enable traps snmp authentication

Enables SNMP server to send authentication failure notifications.


How to Enable the SNMP Server to Send All Enterprise Notifications

At the SCE(config)# prompt, type:

Command
Purpose

snmp-server enable traps enterprise

Enables SNMP server to send all enterprise notifications.


How to Enable the SNMP Server to Send a specific Enterprise Notification

At the SCE(config)# prompt, type:

Command
Purpose

snmp-server enable traps enterprise [attack|chassis|link-bypass|logger|operational-status|port-operational-status|pull-request-failure|RDR-formatter|session| SNTP|subscriber|system-reset|telnet|vas-traffic-forwarding]

Specify the desired enterprise trap type.


Enabling the SNMP Server to Send a Specific Enterprise Notification: Example

The following example shows how to configure the SNMP server to send the logger enterprise notification only.

SCE(config)#snmp-server enable traps enterprise logger

How to Restore All Notifications to the Default Status

At the SCE(config)# prompt, type:

Command
Purpose

default snmp-server enable traps

Resets all notifications supported by the SCE platform to their default status.