Chapter 1 RRPP Configuration
The rapid ring protection protocol (RRPP)
is a link layer protocol designed for Ethernet rings. RRPP can prevent broadcast
storm caused by data loops when the Ethernet rings are healthy, and restore
rapidly the communication paths between nodes after a link is disconnected on
the Ethernet ring network.
Compared with the Spanning Tree Protocol (STP),
RRPP has the following characteristics:
l
Dedicated to Ethernet ring topology
l
Fast response
Figure 1-1 RRPP networking
1. Domain
A domain consists of switches with the same
domain ID and control VLAN. A domain can consist of multiple Ethernet rings,
one of which is the primary ring and the others are subrings. The ring roles
are determined by user configuration.
As shown in Figure
1-1, Domain 1 is an RRPP domain, which consists of Ethernet ring 1 and ring
2. All the nodes on the Ethernet rings belong to the RRPP domain.
2. Ethernet ring
An Ethernet ring is a ring-shaped Ethernet
topology, on which a RRPP domain is based. An RRPP domain consists of a primary
ring and one or more subrings. In configuration, the level of the primary ring
is level 0, and that of the subrings is level 1.
As shown in Figure
1-1, RRPP domain 1 consists of ring 1 and ring 2. If their levels are set
to level 0 and level 1 respectively, ring 1 is the primary ring and ring 2 is
the subring.
Each ring is in one of the following two
states:
l
Healthy state: The physical links of the ring
network are connected.
l
Broken state: A certain physical link is
disconnected on the ring network.
3. Control VLAN and data VLAN
l
A control VLAN is a special VLAN used to
transfer RRPP packets. The port on each switch for connecting the switch with
the Ethernet ring belongs to the control VLAN, and only the ports connected to
the Ethernet ring can be added to the control VLAN. It is not allowed to
configure an IP address for the interface of the control VLAN. You can
configure the control VLAN of the primary ring. The control VLAN of the subring
is assigned by the system automatically. The ID of the subring control VLAN is
the ID of the primary ring control VLAN plus 1.
l
A data VLAN is used to transfer data packets. A
data VLAN contains the ports connecting the switch with the Ethernet ring
network and other ports.
4. Node
Every switch on an Ethernet ring network is
a node. Node roles are as follows:
l
Master node: The node that initiates loop
detection and prevents data loops prevention is the master node. Each ring has
one and only one master node.
l
Transit node: All nodes other than the master
node on a ring are transit nodes.
l
Edge node: An edge node is located on the
primary ring and a subring at the same time. An edge node serves as a transit
node on the primary ring and an edge node on a subring. In an RRPP domain,
there are two edge nodes on a subring. You must specify one of them as
assistant edge node.
The node roles are determined by user
configuration. As shown in Figure 1-1, Switch A is
the master node on ring 1, and Switch B, Switch C and Switch D are transit
nodes on ring 1. Switch B and Switch C are edge nodes because they are both on
ring 2. You can specify one of them as edge node, and the other as assistant
edge node.
5. Primary port and secondary port
The master node and each of the transit
nodes are connected to an Ethernet ring through two ports, of which one is the
primary port and the other is the secondary port. The node roles are determined
by user configuration.
l
The primary port and secondary port of the
master node
The primary port of the master node
transmits the loop detection packet, and the secondary port of the master node
receives the loop detection packet.
When an Ethernet ring is in the healthy
state, the secondary port of the master node allows only RRPP packets to pass,
but logically blocks data packets in data VLANs.
When the Ethernet ring is in the broken
state, the secondary port of the master node stops blocking the data VLAN and
begins to forwards data packets in data VLANs.
l
The primary and secondary ports of a transit
node are functionally the same.
The node roles are determined by user
configuration. As shown in Figure 1-1, Switch A is
the master node of Ring 1. Port 1 and Port 2 of Switch A are the primary port
and secondary port respectively. Switch B, Switch C and Switch D are the
transit nodes on ring 1, and their respective port 1 and port 2 are the primary
port and secondary port on ring 1.
6. Common port and edge port
Of the two ports connecting an edge node to
a subring, one is the common port and the other is the edge port of the node.
The common port connects the edge node to the primary ring and a subring at the
same time. An edge port is connected only with a subring.
The node roles are determined by user
configuration. As shown in Figure 1-1, Switch B
and Switch C are on ring 1 and ring 2 at the same time. Port 2 of Switch B and
Port 1 of Switch C connect the primary and a subring, so they are common ports.
Port 3 of Switch B and Port 3 of Switch C connect only the subring, so they are
edge ports.
7. MAC address FDB
The Layer 2 forwarding database (FDB) on a
switch is updated through the source MAC address auto-learning function of the
switch.
8. Timer
Two timers, Hello timer and Fail timer, are
involved when the master node sends and receives RRPP packets.
l
Hello timer: Defines the time interval at which
the primary port of the master node sends the health detection packet.
l
Fail timer: Defines the timeout time for the
secondary port of the master node to receive health detection packets. The
value of Fail timer must be greater than or equal to three times the Hello
timer value.
The following table describes RRPP packet
types.
Table 1-1
RRPP packets
|
Packet type
|
Description
|
|
Health detection packet
|
The master node sends the health
detection packet (HELLO packet) to detect whether the ring network is
complete.
|
|
LINK UP packet
|
A transit node sends this packet to
notify the master node that a link is UP on the ring network.
|
|
LINK DOWN packets
|
A transit node sends this packet to
notify the master node that a port is DOWN and the ring is broken.
|
|
Common Flush packet
|
The master node sends this packet to tell
all the transit nodes to refresh their respective MAC address FDB.
|
|
Complete Flush packet
|
The master node sends this packet to tell
all the transit nodes to refresh their respective MAC address FDB and unblock
the ports in the blocked data VLANs.
|
I. Link DOWN notification mechanism
When detecting a port in the RRPP domain is
down, a transit node sends the LINK DOWN packet immediately to the master node.
After receiving the LINK DOWN packet, the master node unblocks the data VLAN of
the secondary port, and sends the Common Flush packet to tell all transit nodes
to refresh their respective MAC address FDB.
II. Polling mechanism
The primary port of the master node
periodically sends the health detection packet in a control VLAN.
l
If the secondary port of the master node
receives the health detection packet, this indicates that the ring link is
complete, and the master node will keep the secondary port blocked.
l
If the secondary port of the master node fails
to receive the health detection packet within the predefined timeout time, this
indicates that a failure has occurred to the ring link. In this case, the
master node unblocks the data VLANs on the secondary port, and sends the Common
Flush packet to tell all transit nodes to refresh their respective MAC address
FDB.
III. Ring recovery
The master node may detect that the ring
has recovered a period time after the RRPP domain port on a transit node becomes
UP again. In this period, a temporary data loop may occur in data VLANs, which
can cause broadcast storm.
To avoid temporary data loops, when
detecting the port through which it connects to the ring network becomes UP
again, a transit node blocks the port temporarily (only control VLAN packets
are permitted to pass), and keeps the port blocked until it receives the
Complete Flush packet from the master node.
To ensure normal RRPP operation, you must
configure RRPP correctly. Here are several typical networking applications.
1. Single ring network
Figure
1-2 Single ring network
There is only one ring in the network
topology. In this case, only one RRPP domain is to be defined.
2. Tangent ring networking
Figure 1-3 Tangent ring networking
There are two or more rings in the network
topology and only one common node exists between each pair of rings. In this
case, one RRPP domain must be defined for each ring.
3. Intersectant ring networking
Figure
1-4 Intersectant ring networking
There are two or more rings in the network topology and two common
nodes exist between each pair of rings. In this case, only one RRPP domain is
to be defined, in which one ring must be defined as the primary ring and the
rest as subrings.
The switch ports connecting the Ethernet
rings have been configured as trunk ports. All ports allow data VLAN packets to
pass.
The following table describes the master
node configuration tasks.
Table 1-2
Configure the master node
|
Operation
|
Command
|
Description
|
|
Enter system view
|
system-view
|
-
|
|
Create an RRPP domain, and enter RRPP
domain view
|
rrpp domain domain-id
|
Required.
The command prompt of RRPP domain view
depends on the domain-id you input.
|
|
Specify a control VLAN for the RRPP
domain
|
control-vlan vlan-id
|
Required
|
|
Specify the current switch as the primary
node of a ring, and specify the primary port and the secondary port of the
node
|
ring ring-id node-mode master
[ primary-port pri-port ] ] secondary-port sec-port
] level level-value
|
Required.
Level 0 identifies the primary ring and
level 1 identifies a subring.
|
|
Configure RRPP domain timers
|
timer hello-timer hello-value
fail-timer fail-value
|
Optional.
By default, the Hello timer is set to 1
second, and the Fail timer to 3 seconds.
|
|
Enable an RRPP ring
|
ring ring-id enable
|
Required
|
|
Return to system view
|
quit
|
-
|
|
Enable the RRPP protocol
|
rrpp enable
|
Required
|
|
Display the brief information of all RRPP
domains configured on the switch
|
display rrpp brief
|
Optional.
You can execute the display
command in any view
|
|
Display RRPP configuration details on the
switch
|
display rrpp verbose domain domain-id
[ ring ring-id ]
|
|
Display RRPP packet statistics of the
switch
|
display rrpp statistics domain domain-id
[ ring ring-id ]
|
To clear the RRPP statistics information,
use the reset rrpp statistics domain domain-id [ ring
ring-id ] command.
Caution:
l
The control VLAN for an RRPP domain must not be
a VLAN you have created on the switch, and you are not recommended to configure
the same VLAN as both control VLAN and remote-probe VLAN.
l
Before creating an RRPP ring, you must create a
control VLAN.
l
When deleting an RRPP domain by using the undo
rrpp domain command, make sure no RRPP ring exists in the RRPP domain.
I. Network requirements
l
Define the switch as a node in RRPP domain 1
l
Define VLAN 4092 as the control VLAN
l
Define the switch as the master node on primary
ring 1 in RRPP domain 1, Ethernet1/0/1 as the primary port, and Ethernet1/0/2
as the secondary port.
l
Set the Hello timer and Fail time to 2 seconds
and 7 seconds respectively.
II. Configuration procedure
Caution:
Make sure that the switch ports connecting the Ethernet rings have been configured
as trunk ports. All ports allow data VLAN packets to pass.
<H3C> system-view
[H3C] rrpp domain 1
[H3C-rrpp-domain-1] control-vlan 4092
[H3C-rrpp-domain-1] ring 1 node-mode master
primary-port Ethernet1/0/1 secondary-port Ethernet1/0/2 level 0
[H3C-rrpp-domain-1] timer hello-timer
2 fail-timer 7
[H3C-rrpp-domain-1] ring 1 enable
[H3C-rrpp-domain-1] quit
[H3C] rrpp
enable
[H3C] display
rrpp brief
[H3C] display rrpp verbose domain 1
[H3C] display
rrpp statistics domain 1
The switch ports connecting the Ethernet
rings have been configured as trunk ports. All ports allow data VLAN packets to
pass.
The following table describes the transit
node configuration tasks.
Table 1-3
Configure a transit node
|
Operation
|
Command
|
Description
|
|
Enter system view
|
system-view
|
-
|
|
Create an RRPP domain, and enter RRPP
domain view
|
rrpp domain domain-id
|
Required.
The command prompt of RRPP domain view
depends on the domain-id you input.
|
|
Specify a control VLAN for the RRPP
domain
|
control-vlan vlan-id
|
Required
|
|
Specify the current switch as the transit
node of a ring, and specify the primary port and the secondary port of the
node
|
ring ring-id node-mode transit
[ primary-port pri-port ] [ secondary-port sec-port
] level level-value
|
Required.
Level 0 identifies the primary ring and
level 1 identifies a subring.
|
|
Enable an RRPP ring
|
ring ring-id enable
|
Required
|
|
Return to system view
|
quit
|
-
|
|
Enable RRPP
|
rrpp enable
|
Required
|
|
Display the brief information of all RRPP
domains configured on the switch
|
display rrpp brief
|
Optional.
You can execute the display
command in any view
|
|
Display RRPP configuration details on the
switch
|
display rrpp verbose domain domain-id
[ ring ring-id ]
|
|
Display the RRPP packet statistics on the
switch
|
display rrpp statistics domain domain-id
[ ring ring-id ]
|
To clear the RRPP statistics information,
use the reset rrpp statistics domain domain-id [ ring
ring-id ] command.
Caution:
l
The control VLAN for the RRPP domain must not be
a VLAN you have created on the switch, and you are not recommended to configure
the same VLAN as both control VLAN and remote-probe VLAN.
l
Before creating an RRPP ring, you must create a
control VLAN.
l
When deleting an RRPP domain by using the undo
rrpp domain command, make sure no RRPP ring exists in the RRPP domain.
I. Network requirements
l
Define the switch as a node in RRPP domain 1.
l
Define VLAN 4092 as the control VLAN
l
Define the switch as a transit node on primary
ring 1 in RRPP domain 1, Ethernet1/0/1 as the primary port, and Ethernet1/0/2
as the secondary port.
II. Configuration procedure
Caution:
Make sure that the switch ports connecting the Ethernet rings have been configured
as trunk ports. All ports allow data VLAN packets to pass.
<H3C> system-view
[H3C] rrpp domain 1
[H3C-rrpp-domain-1] control-vlan 4092
[H3C-rrpp-domain-1] ring 1 node-mode transit
primary-port Ethernet1/0/1 secondary-port Ethernet1/0/2 level 0
[H3C-rrpp-domain-1] ring 1 enable
[H3C-rrpp-domain-1] quit
[H3C] rrpp
enable
[H3C] display
rrpp brief
[H3C] display rrpp verbose domain 1
[H3C] display rrpp statistics domain
1
The switch ports connecting the Ethernet
rings have been configured as trunk ports. All ports allow data VLAN packets to
pass.
The following table describes the edge node
configuration tasks.
Table 1-4
Configure an edge node
|
Operation
|
Command
|
Description
|
|
Enter system view
|
system-view
|
-
|
|
Create an RRPP domain, and enter RRPP
domain view
|
rrpp domain domain-id
|
Required.
The command prompt of RRPP domain view depends
on the domain-id you input.
|
|
Specify a control VLAN for the RRPP
domain
|
control-vlan vlan-id
|
Required
|
|
Specify the current switch as a transit
node of the primary ring, and specify the primary port and the secondary
port
|
ring ring-id node-mode transit
[ primary-port pri-port ] [ secondary-port sec-port
] level level-value
|
Required.
Level 0 identifies the primary ring and
level 1 identifies a subring.
|
|
Specify the current switch as an edge node
of the subring, and specify a common port and an edge port
|
ring ring-id node-mode edge
[ common-port comm-port ] [ edge-port
edge-port ]
|
Required
|
|
Enable the primary ring
|
ring ring-id enable
|
Required
|
|
Enable the subring
|
ring ring-id enable
|
Required
|
|
Return to system view
|
quit
|
-
|
|
Enable RRPP
|
rrpp enable
|
Required
|
|
Display the brief information of all RRPP
domains configured on the switch
|
display rrpp brief
|
Optional.
You can execute the display
command in any view
|
|
Display RRPP configuration details on the
switch
|
display rrpp verbose domain domain-id [ ring ring-id ]
|
|
Display the RRPP packet statistics on the
switch
|
display rrpp statistics domain domain-id [ ring ring-id ]
|
To clear the RRPP statistics information,
use the reset rrpp statistics domain domain-id [ ring
ring-id ] command.
Caution:
l
The control VLAN for the RRPP domain must not be
a VLAN you have created on the switch, and you are not recommended to configure
the same VLAN as both control VLAN and remote-probe VLAN.
l
Before creating an RRPP ring, you must create a
control VLAN.
l
When deleting an RRPP domain by using the undo
rrpp domain command, make sure no RRPP ring exists in the RRPP domain.
l
In the same RRPP domain, different rings must
not have the same Ring ID.
I. Network requirements
l
Define the switch as a node in RRPP domain 1.
l
Define
