HP 5500 Ei 5500 Si Switch Series Configuration Guide
Have a look at the manual HP 5500 Ei 5500 Si Switch Series Configuration Guide online for free. It’s possible to download the document as PDF or print. UserManuals.tech offer 1114 HP manuals and user’s guides for free. Share the user manual or guide on Facebook, Twitter or Google+.
![Page 2475
174
Step Command Remarks
5. Establish a cluster.
• Manually establish a cluster:
build cluster-name
• Automatically establish a
cluster:
auto-build [ recover ] Use either approach.
By default, the switch is not the
management switch.
Handshake packets use UDP port 40000. For a cluster to
be established successfully, make sure that the
port is not in use before establishing it.
Enabling management VLAN auto-negotiation
The management VLAN limits the cluster manageme nt range. If...](http://img.usermanuals.tech/thumb/36/58909/w64_hp-5500-ei-5500-si-switch-series-configuration-guide-2474.png "Page 2475
174
Step Command Remarks
5. Establish a cluster.
• Manually establish a cluster:
build cluster-name
• Automatically establish a
cluster:
auto-build [ recover ] Use either approach.
By default, the switch is not the
management switch.
Handshake packets use UDP port 40000. For a cluster to
be established successfully, make sure that the
port is not in use before establishing it.
Enabling management VLAN auto-negotiation
The management VLAN limits the cluster manageme nt range. If...")
170 Task Remarks Manually collecting topology information Optional Enabling the cluster function Optional Establishing a cluster Required Enabling management VLAN auto-negotiation Required Configuring communication between the management switch and the member switches within a cluster Optional Configuring cluster management protocol packets Optional Cluster member management Optional Configuring the member switches: Enabling NDP Optional Enabling NTDP Optional Manually collecting topology information Optional Enabling the cluster function Optional Deleting a member switch from a cluster Optional Configuring access between the management switch and its member switches Optional Adding a candidate switch to a cluster Optional Configuring advanced cluster management functions: Configuring topology management Optional Configuring interaction for a cluster Optional SNMP configuration synchronization function Optional Configuring web user accounts in batches Optional Configuring the management switch Perform the tasks in this section to configure the management switch for a cluster. Enabling NDP globally and for specific ports For NDP to work normally, you must enable NTDP both globally and on specific ports. To enable NDP globally and for specific ports: Step Command Remarks 1. Enter system view. system-view N/A 2. Enable NDP globally. ndp enable Optional. By default, this function is enabled.
171 Step Command Remarks 3. Enable the NDP feature on ports. • In system view: ndp enable interfaceinterface-list • In Ethernet interface view or Layer 2 aggregate interface view: a. interface interface-type interface-number b. ndp enable Use either command. By default, NDP is enabled globally and also on all ports. HP recommends that you disable NDP on a port which connects with the switches that do not need to join the cluster. This prevents the management switch fr om adding and collecting topology information from switches which do not need to join the cluster. Configuring NDP parameters A port enabled with NDP periodically sends NDP packets to its neighbors. If no NDP information from the neighbor is received to reset the holdtime, the holdtime times out and the switch removes the corresponding entry from the NDP table. The time for the receiving switch to hold NDP packets cannot be shorter than the interval for sending NDP packets. Otherwise, the NDP table may become instable. To configure NDP parameters: Step Command Remarks 1. Enter system view. system-view N/A 2. Configure the interval for sending NDP packets. ndp timer hello hello-time Optional. The default interval is 60 seconds. 3. Configure the period for the receiving switch to keep the NDP packets. ndp timer aging aging-time Optional. The default setting is 180 seconds. Enabling NTDP globally and for specific ports For NTDP to work normally, you must enable NTDP both globally and on specific ports. To enable NTDP globally and for specific ports: Step Command Remarks 1. Enter system view. system-view N/A 2. Enable NTDP globally. ntdp enable Optional. By default, NTDP is enabled globally. 3. Enter Ethernet interface view or Layer 2 aggregate interface view. interface interface-type interface-number N/A
172 Step Command Remarks 4. Enable NTDP for the port. ntdp enable Optional. By default, NTDP is enabled on all ports. HP recommends that you disable NTDP on a port which connects with the switches that do not need to join the cluster. This prevents the management sw itch from adding and collecting topology information from switches which do not need to join the cluster. Configuring NTDP parameters By configuring the maximum hops for collecting topo logy information, you can get topology information of the switches in a specified range, thus avoiding unlimited topology collection. After the interval for collecting topology information is configured, the switch collects the topology information at this interval. To avoid network congestion caused by large amounts of topology responses received in short periods: • Upon receiving an NTDP topology-collection request, a switch does not forward it immediately. Instead, it waits for a period of time and then forwards the NTDP topology-collection request on its first NTDP-enabled port. • Except for its first port, each switch’s NTDP-enabled ports wait for a period of time, and then forward the NTDP topology collection request after the previous port forwards it. To configure NTDP parameters: Step Command Remarks 1. Enter system view. system-view N/A 2. Configure the maximum hops for topology collection. ntdp hop hop-value Optional. The default setting is 3. 3. Configure the interval for collecting topology information. ntdp timer interval Optional. The default interval is 1 minute. 4. Configure the delay to forward topology-collection request packets on the first port. ntdp timer hop-delay delay-time Optional. The default setting is 200 ms. 5. Configure the port delay to forward topology-collection request on other ports. ntdp timer port-delay delay-time Optional. The default setting is 20 ms. The two delay values should be configured on the topology collecting switch. A topology-collection request sent by the topology collecting switch carrie s the two delay values, and a switch that receives the request forwards the request according to the delays. Manually collecting topology information The management switch collects topology information periodically after a cluster is created. In addition, you can manually start topology information collect ion on the management switch or NTDP-enabled
173
switch, thus managing and monitoring switches in real time, regardless of whether a cluster is created.
To configure to manually collect topology information:
Task Command
Manually collect topology information. ntdp explore
Enabling the cluster function
Step Command Remarks
1. Enter system view.
system-view N/A
2. Enable the cluster function
globally. cluster enable Optional.
By default, this function is enabled.
Establishing a cluster
Before establishing a cluster, you need to specif
y the management VLAN, and you cannot modify the
management VLAN after a device is added to the cluster.
In addition, you need to configure a private IP address pool for the devices to be added to the cluster on
the device to be configured as the management device before establishing a cluster. Meanwhile, the IP
addresses of the VLAN interfaces of the management device and member devices cannot be in the same
network segment as that of the cluster address pool; otherwise, the cluster cannot work normally. When
a candidate device is added to a cluster, the management device assigns it a private IP address for it to
communicate with other devices in the cluster.
You can establish a cluster in two ways: manually and automatically. You can follow the prompts to
establish a cluster automatically. The system cluster auto-establishment process will prompt you through
the following steps:
1. Enter a name for the cluster you want to establish.
2. List all the candidate switches within your predefined hop count.
3. Start to automatically add them to the cluster.
Yo u c a n p r e s s Ctrl+C anytime during the adding process to exit the cluster auto-establishment process.
However, this will only stop adding new switches in to the cluster, and switches already added into the
cluster are not removed.
To establish a cluster manually or automatically:
Step Command Remarks
1. Enter system view.
system-view N/A
2. Specify the management VLAN. management-vlan vlan-id Optional.
By default, VLAN 1 is the
management VLAN.
3.
Enter cluster view.
cluster N/A
4. Configure the private IP address
range for member switches. ip-pool
ip-address { mask |
mask-length } By default, the private IP address
range for member switches is not
configured.
174 Step Command Remarks 5. Establish a cluster. • Manually establish a cluster: build cluster-name • Automatically establish a cluster: auto-build [ recover ] Use either approach. By default, the switch is not the management switch. Handshake packets use UDP port 40000. For a cluster to be established successfully, make sure that the port is not in use before establishing it. Enabling management VLAN auto-negotiation The management VLAN limits the cluster manageme nt range. If the switch discovered by the management switch does not belong to the mana gement VLAN, meaning the cascade ports and the ports connecting with the management switch do no t allow the packets from the management VLAN to pass, and the new switch cannot be added to the cluster. Through the configuration of the management VLAN auto-negotiation function, the cascade ports an d the ports directly connected to the management switch can be automatically added to the management VLAN. Configuration guidelines When the management VLAN auto-negotiation is enabled, the ports connecting member switches change as follows: • If a port was an access port, after changing to a hybrid port, the port does not permit the packets of any other VLAN to pass through except the management VLAN, which passes as tagged. • If a port was a trunk or a hybrid port, the link type change process does not affect the port type and the permitted VLANs. The only change is permit the packets of the management VLAN to pass through tagged only for a hybrid port. Before enabling this function, check the link type s of ports connecting member switches and the VLANs whose packets are permitted to pass through to avoid influence to your network due to link type change of ports. To configure management VLAN auto-negotiation: Step Command Remarks 1. Enter system view. system-view N/A 2. Enter cluster view. cluster N/A 3. Enable management VLAN auto-negotiation. management-vlan synchronization enable By default, this function is disabled. Configuring communication between the management switch and the member switches within a cluster In a cluster, the management switch and member switches communicate by sending handshake packets to maintain connection between them. You can config ure interval of sending handshake packets and the holdtime of a switch on the management switch. This configuration applies to all member switches within the cluster. For a member switch in Connect state: • If the management switch does not receive hand shake packets from a member switch within the
175 holdtime, it changes the state of the member switch to Disconnect. When the communication is recovered, the member switch needs to be re-added to the cluster (this process is automatically performed). • If the management switch receives handshake packe ts from the member switch within the holdtime, the state of the member switch remains Active and the holdtime is restarted. To configure communication between the management sw itch and the member switches within a cluster: Step Command Remarks 1. Enter system view. system-view N/A 2. Enter cluster view. cluster N/A 3. Configure the handshake interval. timer interval Optional. The default interval is 10 seconds. 4. Configure the holdtime of a switch. holdtime hold-time Optional. The default setting is 60 seconds. Configuring cluster management protocol packets By default, the destination MAC address of cluster management protocol packets (including NDP, NTDP and HABP packets) is a multicast MAC address 0180-C200-000A, which IEEE reserved for later use. Since some switches cannot forward the multicast packets with the destination MAC address of 0180 -C200 -000A, so cluster management packets cannot traverse these switches. For a cluster to work normally in this case, you can modify the destination MAC address of a cluster management protocol packet without changing the current networking. The management switch periodically sends MAC address negotiation broadcast packets to advertise the destination MAC address of the cluster management protocol packets. When you configure the destination MAC address for cluster management protocol packets: • If the interval for sending MAC address negoti ation broadcast packets is 0, the system automatically sets it to 1 minute. • If the interval for sending MAC address negotiation broadcast packets is not 0, the interval remains unchanged. To configure the destination MAC address of the cluster management protocol packets: Step Command Remarks 1. Enter system view. system-view N/A 2. Enter cluster view. cluster N/A
176
Step Command Remarks
3. Configure the
destination MAC
address for cluster
management protocol
packets. cluster-mac
mac-address By default, the destination MAC address is
0180-C200-000A.
The following are the configurable MAC
addresses:
•
0180-C200-0000
• 0180-C200-000A
• 0180-C200-0020 through
0180-C200-002F
• 010F-E200-0002
4. Configure the interval for
sending MAC address
negotiation broadcast
packets. cluster-mac
syn-interval
interval Optional.
The default interval is one minute.
Cluster member management
You can manually add a candidate switch to a cluster, or remove a member switch from a cluster.
If a member switch needs to be rebooted for so
ftware upgrade or configuration update, you can
remotely reboot it through the management switch.
Adding a member switch
Step Command
1. Enter system view.
system-view
2. Enter cluster view.
cluster
3. Add a candidate switch to the cluster. add-member
[ member-number ] mac-address mac-address
[ password password ]
Removing a member switch
Step Command
1. Enter system view.
system-view
2. Enter cluster view.
cluster
3. Remove a member switch from the cluster.
delete-member member-number [ to-black-list ]
Rebooting a member switch
Step Command
1. Enter system view.
system-view
2. Enter cluster view.
cluster
3. Reboot a specified member switch. reboot member
{ member-number | mac-address
mac-address } [ eraseflash ]
177 Configuring the member switches Enabling NDP See Enabling NDP globally and for specific ports . Enabling NTDP See Enabling NTDP globally and for specific ports . Manually collecting topology information See Manually collecting topology information . Enabling the cluster function See Enabling the cluster function . Deleting a member switch from a cluster Step Command 1. Enter system view. system-view 2. Enter cluster view. cluster 3. Delete a member switch from the cluster. undo administrator-address Configuring access between the management switch and its member switches After having successfully configured NDP, NTDP and cluster, you can configure, manage and monitor the member switches through the management switch. You can manage member switches in a cluster through switching from the operation interface of the management switch to that of a member switch or configure the management switch by switching from the operation interface of a member switch to that of the management switch. Configuration guidelines Telnet connection is used in the switching between the management switch and a member switch. Note the following when switching between them: • Authentication is required when you switch from a member switch to the management switch. The switching fails if authentication is not passed. If authentication is passed, your user level is allocated by the management switch according to the predefined level. • When a candidate switch is added to a cluster and becomes a member switch, its super password with the level of 3 will be automatically synchronized to the management switch. Therefore, after a cluster is established, it is not recommended to modify the super password of any member
178
(including the management switch and member switches) of the cluster; otherwise, the switching
may fail because of an authentication failure.
• If the member specified in this command does not exist, the system prompts error when you execute
the command; if the switching succeeds, your user level on the management switch is retained.
• If the Telnet users on the switch to be logged in reach the maximum number, the switching fails.
• To prevent resource waste, avoid ring switching when configuring access between cluster members.
For example, if you switch from the operation in terface of the management switch to that of a
member switch and then need to switch back to that of the management switch, use the quit
command to end the switching, but not the cluster switch-to administrator c o m m a n d t o s w i t c h t o t h e
operation interface of the management switch.
Configuration procedure
To configure access between member switches of a cluster:
Step Command
1. Switch from the operation interface of the management
switch to that of a member switch. cluster switch-to { member-number
|
mac-address mac-address | sysname
member-sysname }
2. Switch from the operation interface of a member switch to
that of the management switch. cluster switch-to
administrator
Adding a candidate switch to a cluster
Step Command
1. Enter system view.
system-view
2. Enter cluster view.
cluster
3. Add a candidate switch to the cluster.
administrator-address mac-address name name
Configuring advanced cluster management
functions
Configuring topology management
The concepts of blacklist and whitelist are used for topology management. An administrator can
diagnose the network by comparing the current topo logy (namely, the information of a node and its
neighbors in the cluster) and the standard topology.
• Topology management whitelist (standard topology): A whitelist is a list of topology information
that has been confirmed by the administrator as correct. You can get the information of a node and
its neighbors from the current topology. Based on the information, you can manage and maintain
the whitelist by adding, dele ting or modifying a node.
• Topology management blacklist: Switches in a blackl ist are not allowed to join a cluster. A blacklist
contains the MAC addresses of switches. If a blackl isted switch is connected to a network through
179
another switch not included in the blacklist, the MAC address and access port of the latter are also
included in the blacklist. The cand idate switches in a blacklist can be added to a cluster only if the
administrator manually removes them from the list.
The whitelist and blacklist are mutually exclusive. A whitelist member cannot be a blacklist member, and
the blacklist member cannot be a whitelist member. Ho wever, a topology node can belong to neither the
whitelist nor the blacklist. Nodes of this type are usually newly added nodes, whose identities are to be
confirmed by the administrator.
You can back up and restore the whitelist and blacklist in the following two ways:
• Backing them up on the FTP server shared by the cluster. You can manually restore the whitelist and
blacklist from the FTP server.
• Backing them up in the Flash of the management switch. When the management switch restarts, the
whitelist and blacklist will be automatically restored from the Flash. When a cluster is re-established,
you can choose whether to restore the whitelist and blacklist from the Flash automatically, or you
can manually restore them from the Flash of the management switch.
To configure cluster topology management:
Step Command Remarks
1. Enter system view.
system-view N/A
2. Enter cluster view.
cluster N/A
3. Add a switch to the blacklist.
black-list add-mac mac-address Optional.
4. Remove a switch from the
blacklist. black-list delete-mac
{ all | mac-address } Optional.
5. Confirm the current topology
and save it as the standard
topology. topology accept
{ all [ save-to { ftp-server |
local-flash } ] | mac-address mac-address |
member-id member-number } Optional.
6.
Save the standard topology to
the FTP server or the local
Flash. topology save-to {
ftp-server | local-flash } Optional.
7. Restore the standard topology
information. topology restore-from
{ ftp-server | local-flash } Optional.
Configuring interaction for a cluster
After establishing a cluster, you can configure FTP/TF TP server, NM host and log host for the cluster on
the management switch.
• After you configure an FTP/TFTP server for a cluster, the members in the cluster access the FTP/TFTP
server configured through the management switch. Execute the ftp server-address or tftp
server-address command and specifying the private IP address of the management switch as the
server-address . For more information about the ftp and tftp commands, see Fundamentals
Command Reference .
• After you configure a log host for a cluster, all the log information of the members in the cluster will
be output to the configured log host in the following way:
{ Member switches send their log information to the management switch.
{ The management switch converts the addresses of log information and sends them to the log
host.