Configuring ERPS

Summary: Networking › Switching › Edge › Synapse

Overview of ERPS

ERPS (Ethernet Ring Protection Switching) was developed by ITU, also known as G.8032.It is a link layer protocol specifically applied to Ethernet. It can prevent the broadcast storm caused by the data loop when the Ethernet ring network is complete, and can quickly restore the communication between each node on the ring network when a link on the Ethernet ring is disconnected. At present, the technology to solve the Layer 2 network loop problem is STP.STP is more mature to use, but its convergence time is longer (seconds).ERPS is a link layer protocol that is specially applied to Ethernet and has a faster rate than STP for convergence, up to 50ms. ERPS typical scenario：

Introduction to ERPS Rationale

ERPS is a standard ring network protocol dedicated to the Ethernet link layer, with the ERPS ring as the basic unit. Only two ports on each layer 2 switch can be added to the same ERPS ring. In the ERPS, in order to prevent network loop, a break-down mechanism can be launched, blocking the RPL owner port and eliminating the ring route. When the ring connection fails, the equipment running the ERPS protocol can quickly forward the blocked port, make the link protection replacement, and restore link communication between various nodes on the ring network. This section mainly presents the rationale for the implementation of ERPS under the basic network based on the normal ->link failure->link recovery process (including protection switch operations).

Link OK

As shown in the diagram below, the equipment on the ring consisting of SwitchA~SwitchE is in good condition. To prevent loops, ERPS first blocks the RPL owner port. If the RPL neighbor port is configured, the port will also be blocked, and other ports can forward traffic normally.

Link Failure

As shown in the diagram, when the link between SwitchD and SwitchE fails, the ERPS protocol starts the protection switching mechanism, blocks the ports on both ends of the faulty link, and then forward the RPL owner port, and the two ports resume user traffic. receiving and sending, thus ensuring uninterrupted traffic.

Link Restore

After the link returns to normal, if the ERPS ring is configured in revert mode, the device where the RPL owner port resides will block the traffic on the RPL link again, and the faulty link will be used again to transmit user traffic.

Configuring

• Creating Ring

SWITCH(config)#erps ring <1-255> east-interface IFNAME west-interface IFNAME

SWITCH(config)#no erps ring <1-255>

Create/delete ERPS ring. The ERPS ring is made up of the same set of VLAN and interconnected layer 2 switch, which is the basic unit of the ERPS protocol and needs to be configured on each device in the ring. The ring number is the unique identifier for the ERPS ring.

SWITCH(config)#erps instance NAME

SWITCH(config)#no erps instance NAME

Create/remove ERPS instances; Create an instance to go into instance configuration mode. For the layer 2 switch operating an ERPS protocol, VLAN transmitting ERPS and data articles must be mapped into a protective instance so that ERPS protocol can be forwarded or blocked in accordance with their blocking principles. Otherwise, user traffic could cause broadcast storms in a ring network that could make the network unavailable.

SWITCH(config-erps-inst)#ring <1-255>

Configure the corresponding relationships between ERPS instances and rings.

SWITCH(config-erps-inst)#level <0-7>

Configure ERPS instance level.

SWITCH(config-erps-inst)#rpl-role NAME

Configure the ERPS instance RPL role; An ERPS ring has only one RPL owner port, which is determined by user configuration. The RPL owner port is blocked from forwarding user traffic to prevent loops in the ERPS ring.

SWITCH(config-erps-inst)#vlan <1-4094> raps-channel

SWITCH(config-erps-inst)#no raps-channel

Configuration/delete raps VLAN for ERPS instances; Each ERPS ring must be configured with a raps VLAN.Different ERPS rings cannot use the same raps VLAN ID.

SWITCH(config-erps-inst)#protected-mst-instance <0-255>

Configure MST Instance； The relationship between VLAN and Instance can be configured in MST mode, after STP mode be set to MSTP, refer to STP configuration for more details; by default, all VLANs belong to Instance 0; the default value is 0. Note: Multi-instance is currently not supported in intersecting rings!

SWITCH(config-erps-inst)#sub-ring block (east-interface | west-interface)

Configure the ERPS instance as a sub-ring instance and specify a sub-ring block port.

SWITCH(config-erps-inst)#virtual-channel attached-to-instance NAME

SWITCH(config-erps-inst)# non-virtual-channel

Configure the type of ERPS intersecting sub-ring: virtual channel and associated main ring; or non-virtual channel type. Note: The position displayed by this command in show running-config must be after the displayed position of the associated instance. Normally only need to ensure that the sub-ring ID and instance name are larger than the main ring ID and instance name.

SWITCH(config-erps-inst)#revertive | non-revertive

Configure ERPS revertive/non-revertive.

SWITCH(config-erps-inst)# (wtr-timer (<1-12> | default) | holdoff-timer (<0-100> | default) | guard-timer (<1-200> | default))

Configure ERPS timer parameters. <1-12>: in minutes; revert time after recovery, default is 5 minutes. <0-100>：in 100 milliseconds; hold time before port forwarding, the default is 0, direct forwarding without delay. <1-200>：in 10 milliseconds; protection window when state changes, avoid receiving messages from previous state leading to protocol errors, default is 50：500 ms. guard-timer parameters limit network size. It is conservatively recommended that when there are more than 300 nodes in the ring network, directly configure this parameter to the maximum value to avoid the failure of old packets to be discarded due to the large network size; no special configuration is required for nodes within 300 nodes.

SWITCH(config)# erps logging

SWITCH(config)# no erps logging

Configure ERPS logging.

Examples

1. Single-ring case requirements: As shown in the figure, the configuration blocks the direct links of S1 and S2 by default, and restores the link in time to ensure the availability of the network in case of failure. Where the data VLANs are 1, 2 and 3. S1/S2：

• Enter global configuration mode, create ERPS and set related parameters, command reference list below：

Create vlan 2,3;vlan 1 default exists

SWITCH(config)#vlan 2,3

Change the interface mode to trunk. By default, trunk mode will add all data vlans and management vlans to the interface for forwarding.

SWITCH(config)#interface gigabitEthernet0/9-10

SWITCH(config-if)#switchport mode trunk

Create ERPS ring 1

SWITCH(config)#erps ring 1 east gigabitEthernet0/9 west gigabitEthernet0/10

Create ERPS instance 1, associated with ring 1, and associated details configuration

SWITCH(config)#erps instance 1

SWITCH(config-erps-inst)#ring 1

SWITCH(config-erps-inst)#rpl-role non-owner

SWITCH(config-erps-inst)#vlan 1000 raps-channel

S3：

SWITCH(config)#Vlan 2,3

SWITCH(config)#interface gigabitEthernet0/9,gigabitEthernet0/10

SWITCH(config-if)#switchport mode trunk

SWITCH(config)#Erps ring 1 east gigabitEthernet0/9 west gigabitEthernet0/10

SWITCH(config)#Erps instance 1

SWITCH(config-erps-inst)#ring 1

SWITCH(config-erps-inst)#rpl-role owner east

SWITCH(config-erps-inst)#vlan 1000 raps-channel

2. Intersection ring case requirements As shown in the following topology, S1, S2, S3, and S4 form intersecting rings, and the data vlans are 1, 2, 3, and 4. It is required to achieve fast convergence when a single point of failure occurs in each ring; a maximum of two faults can occur in the network Points (different rings), without user disconnection, to achieve optimal reliability. Typical configuration examples： S1: Vlan 2,3,4 interface gigabitEthernet0/9-12 switchport mode trunk Erps ring 1 east gigabitEthernet0/9 west gigabitEthernet0/10 Erps instance 1 ring 1 vlan 1000 raps-channel Erps ring 2 east gigabitEthernet0/9 west gigabitEthernet0/11 Erps instance 2 ring 2 sub-ring block east-interface vlan 1100 raps-channel virtual-channel attached-to-instance 1 S2: Vlan 2,3,4 interface gigabitEthernet0/9-12 switchport mode trunk Erps ring 1 east gigabitEthernet0/9 west gigabitEthernet0/10 Erps instance 1 ring 1 vlan 1000 raps-channel Erps ring 2 east gigabitEthernet0/12 west gigabitEthernet0/10 Erps instance 2 ring 2 sub-ring block east-interface vlan 1100 raps-channel virtual-channel attached-to-instance 1 S3: Vlan 2,3,4 interface gigabitEthernet0/5-6 switchport mode trunk Erps ring 1 east gigabitEthernet0/5 west gigabitEthernet0/6 Erps instance 1 ring 1 rpl-role owner east vlan 1000 raps-channel S4: Vlan 2,3,4 interface gigabitEthernet0/9-12 switchport mode trunk Erps ring 2 east gigabitEthernet0/9 west gigabitEthernet0/10 Erps instance 2 ring 2 rpl-role owner east vlan 1100 raps-channel 3. Tangent ring case requirements The topology diagram is shown below. S1 is located in the central computer room, which can be supervised and maintained by the administrator in real time, and has high reliability; S2-S5 are distributed in various deployment points, in order to improve the reliability of the network and avoid the occurrence of single-link external connection The single-point failure risk is avoided, and the single-machine failure risk that may occur in a dual-link external connection is avoided, and the dual-link external connection is used to form a ring network. It is required that each ring network can converge quickly when a single point of failure occurs to avoid user network interruption. Typical configuration examples： S1: Vlan 2,3,4 interface gigabitEthernet0/9-12 switchport mode trunk Erps ring 1 east gigabitEthernet0/9 west gigabitEthernet0/10 Erps instance 1 ring 1 vlan 1000 raps-channel Erps ring 2 east gigabitEthernet0/11 west gigabitEthernet0/12 Erps instance 2 ring 2 vlan 1100 raps-channel S2: Vlan 2,3,4 interface gigabitEthernet0/9-12 switchport mode trunk Erps ring 1 east gigabitEthernet0/9 west gigabitEthernet0/10 Erps instance 1 ring 1 rpl-role owner east vlan 1000 raps-channel S3: Vlan 2,3,4 interface gigabitEthernet0/5-6 switchport mode trunk Erps ring 1 east gigabitEthernet0/5 west gigabitEthernet0/6 Erps instance 1 ring 1 vlan 1000 raps-channel S4: Vlan 2,3,4 interface gigabitEthernet0/9-12 switchport mode trunk Erps ring 2 east gigabitEthernet0/9 west gigabitEthernet0/10 Erps instance 2 ring 2 rpl-role owner east rpl-role owner east S5: Vlan 2,3,4 interface gigabitEthernet0/9-12 switchport mode trunk Erps ring 2 east gigabitEthernet0/9 west gigabitEthernet0/10 Erps instance 2 ring 2 vlan 1100 raps-channel

Display Information

• Show ERPS Ring Information

SWITCH#show erps ring 1

Ring : 1 ========== Bridge : 1 East : gigabitEthernet0/23 West : gigabitEthernet0/24 ERP Inst :1, 2,

SWITCH#

SWITCH#

SWITCH#show erps instance 1

Name : 1 Protected MST Instance: 0 Protected VLANs : 1 State : ERPS_ST_IDLE Last Priority : RAPS-NR-RB Phy Ring : 1 Role : NON-OWNER East Link : Link_Unblocked(up)(78-A9-12-12-13-12, 1) West Link : Link_Unblocked(up)(78-A9-12-12-13-12, 1) TCN Propagation : Disabled Attached : – Attached To : – Virtual ID : -:- —————————————————– Channel | Interface (LEVL, VID, RID) | (east,ver) , (west,ver) ===================================================== (0, 1000, 1) | (gigabitEthernet0/23, V=1), (gigabitEthernet0/24, V=1) ===================================================== Wait-To-Restore : 5 mins Hold Off Timer : 0 secs Guard Timer : 500 ms Wait-To-Block : 5500 ms Protection Type : Revertive

SWITCH#