Nothing Special   »   [go: up one dir, main page]

2-4-1 RLDP

Download as pdf or txt
Download as pdf or txt
You are on page 1of 10

RLDP Principle

RLDP Overview

Rapid Link Detection Protocol (RLDP) is developed by Ruijie for rapidly detecting Ethernet link
failures.

General Ethernet link detection mechanisms detect link connectivity based on physical connection
status and automatic negotiation at the physical layer. Such mechanisms have limitations and cannot
provide users with reliable link detection information in specific scenarios.
o Scenario 1: Among the optical fiber of port SW1, the core receiving signals fails while the core transmitting signals is normal. As a
fiber converter exists, the peer device displays normal Tx and Rx (the device transmits and receives signals successfully).
o Scenario 2: An Internet service provider (ISP) transmission network exists between two Ethernet devices. If the network fails, the
Ethernet devices at both ends cannot sense.

SW1 SW2

Rx Rx

Tx Tx
RLDP Principle

RLDP defines two kinds of protocol packets: probe packet and echo packet, whose destination
address is L2 multicast address (01-d0-f8-00-00-02).

Each port with RLDP configured and in link up state periodically sends its probe packets and waits for
adjacent ports to respond to the probe packet. It also expects probe packets from adjacent ports. If a
link is normal both physically and logically, the port should be able to receive an echo packet and a
probe packet from the adjacent port. Otherwise, the link is considered abnormal.

Ruijie RLDP provides three functions: loop detection, one-way link detection, and both-way link
detection.

RLDP Probe
RLDP Echo
RLDP Loop Detection Function

RLDP is mainly used on access layer switches for loop detection. (RLDP anti-loop can be used at the distribution
layer as well, but the protection granularity is relatively large.) RLDP is especially applicable to the detection of self-
loop of a downlink hub of a switch, which cannot be prevented by STP.

It is recommended that RLDP be enabled for access switch ports connecting to end users in the project
implementation phase. The optimized configuration that is deployed in advance can prevent loops on the ports.

RLDP loop detection mainly applies to the following two scenarios:

Access Switch

Access Switch
Port enabling portfast

Scenario 1: Switch Self-Loop Scenario 2: Switch Port Self-Loop


Principle of RLDP Loop Detection

If a port of a switch configured with RLDP receives an RLDP packet of the switch itself, a loop occurs on the
port. In this case, RLDP processes the fault according to the user configuration, including generating an
alarm, setting the port in errdisable state, disabling the port SVI, and disabling port learning and forwarding.

Actions in case of a loop:


Switch A
1. block Logically block the port.
2. shutdown-port Set the port in errordisable state.
3. shutdown-svi Disable the port SVI.
4. Warning Produce a warning.

RLDP Probe (Switch A) RLDP Probe (Switch A)


Principle of RLDP One-way Link Detection

If a one-way link failure occurs, the port link can only receive or send packets, which can be caused by, for
example, incorrect connection of optical fibers. As shown in the figure below, if an RLDP port receives only
a probe packet from an adjacent port, a one-way link failure occurs. In this case, RLDP processes the fault
according to the user configuration.

Switch A

RLDP Probe (Switch B) RLDP Probe (Switch A)


RLDP Notes

Routed ports do not support the processing method of shutdown-svi. Therefore, the method can be configured but
will not be executed if a failure is detected on a routed port.

When loop detection is configured, neighbor devices downlink to a port cannot enable RLDP detection. Otherwise,
detection on the port will be incorrect.

If RLDP detects a link failure, a warning is generated. You can configure a log server to record at least three levels
of device logs.

Due to different product features, blocked ports of certain products still send packets to the CPU. As a result, if a
loop is detected on a port and the processing method is configured as block, the port still transmits a large number
of packets to the CPU. This does not achieve the effect of loop detection. Therefore, it is recommended that the
shutdown-port processing method be configured for loop detection.

In RLDP fault processing, the block function and STP are mutually exclusive. If the fault processing method of a
port is configured as block, it is recommended that STP be disabled. Otherwise, if a one-way link failure occurs,
which cannot be detected by STP, STP may allow port forwarding while RLDP blocks the port. To enable RLDP
and STP at the same time, you are advised to configure the fault processing method as shutdown-port.
Principle of RLDP Two-way Link Detection

If a two-way link failure occurs, the port link fails to both receive and transmit packets. As shown in the
figure below, if a device port sends an RLDP probe packet but fails to receive a probe packet and echo
packet from a neighbor, a two-way link failure occurs.

Switch A

RLDP Probe (Switch A)


Typical Application Scenarios of RLDP

[Networking requirements] As shown in the figure below, uses from various departments of an enterprise access
the network through switches A and B. Network loops occur from time to time due to network exceptions caused by
quality issues of optical fibers interconnecting devices or due to incorrect operations of users.

RLDP loop detection and one-way and two-way link detection functions need to be configured to rapidly locate
failures and recover the network in time. Main requirements include:
o 1. If a one-way or two-way failure is detected on the link between switches A and B, shut down the port.

o 2. If a loop is detected on a downlink port of switches A and B, block the port.

o 3. Recover a port five minutes after it is set in errordisable state, and detect all ports that failed the RLDP detection again.

Switch A Gi0/24 Gi0/24 Switch B

Gi0/1-20
Gi0/1-20
Summary

RLDP can detect a one-way link failure, two-way link failure, and loop on a switch port, and
enable different violation actions for detected failures.

You might also like