beautypg.com

Introduction, How ping polling works, Introduction how ping polling works – Allied Telesis AlliedWare Plus Operating System Version 5.4.4C (x310-26FT,x310-26FP,x310-50FT,x310-50FP) User Manual

Page 2052

background image

Ping Polling Introduction and Configuration

Software Reference for x310 Series Switches

78.2

AlliedWare Plus

TM

Operating System - Version 5.4.4C

C613-50046-01 REV A

Introduction

Ping polling lets your device regularly check whether it can reach other hosts on a
network. It works by sending ICMP Echo Requests to a host and waiting for replies sent
back. If ping polling indicates that a host’s status has changed, then your device can
respond to the new status. When a host is unreachable, ping polling continues monitoring
the host’s reachability.

You can configure triggers to activate when ping polling determines that the host’s status
has changed. For example, you could configure a trigger to run a script that opens and
configures an alternative link if the host at the other end of a preferred link becomes
unavailable. You could then configure a second trigger to run a script that automatically
returns traffic to the preferred link as soon as it is available again.

How Ping Polling Works

To determine a host’s reachability, your device regularly sends ICMP Echo Request packets
(“pings”) to the host. As long as your device receives ping responses from the host, it
considers the host to be reachable. If your device does not receive a reply to a set number
of ICMP Echo Requests, it considers that the host is unreachable. It continues to try to ping
the device, at an increased rate. After it receives a set number of responses, it considers the
device to be reachable again.

By default, a polling instance sends a ping every 30 seconds as long as it is receiving
replies. The frequency of this polling is controlled by the

normal-interval

command.

When a reply is not received, the polling instance increases the frequency at which it polls
the device. This frequency is controlled by the

critical-interval

command, and by default,

is set to send a packet every one second. It maintains this higher rate of polling until it has
received sufficient consecutive replies.

The polling instance determines whether a device is reachable or unreachable based on
the settings of the

fail-count

,

sample-size

, and

up-count

commands. To determine

whether a device is reachable, the polling instance counts the number of failed pings
within a set sample size. The sample size is set by the

sample-size

command, and by

default is 5 ping responses. Within the sample size, the number of failed pings that means
that the device is down is set by the

fail-count

command. By default this is set to 5. Once a

polling instance has determined that a device is unreachable, it must receive a set number
of consecutive replies before it changes the device’s status back to reachable. This number
is configured with the

up-count

command.

The following figure illustrates a polling instance where the device becomes unreachable,
then reachable. It uses this configuration:

awplus(config-ping-poll)#

fail-count 4

awplus(config-ping-poll)#

sample-size 5

awplus(config-ping-poll)#

up-count 3

awplus(config-ping-poll)#

critical-interval 1

awplus(config-ping-poll)#

normal-interval 30