Project

General

Profile

Actions

Bug #5215

closed

DHCP failover generates not working configuration.

Added by Christopher Harbort almost 10 years ago. Updated almost 10 years ago.

Status:
Rejected
Priority:
Low
Assignee:
-
Category:
DHCP (IPv4)
Target version:
-
Start date:
09/30/2015
Due date:
% Done:

0%

Estimated time:
Plus Target Version:
Release Notes:
Affected Version:
Affected Architecture:

Description

Using a CARP-Cluster of two nodes with DHCP Server in failover mode does not work, because the generated configuration is wrong.
Setup:

Node1: 172.23.136.131
Node2: 172.23.136.132
CARP: 172.23.136.130

Node2 is configuration master and the dhcpd settings got synchronized to Node1. As long as we do not use the "failover peer IP"-setting, both nodes just run the service without knowing each other. When we set the failover peer IP to Node1, both nodes hang in recover state. This is because the generated configuration(see attached files) for both nodes sets them to secondary and uses the wrong ports, so they can never reach the other node.

I suggest following changes:
Node1(secondary):
.....
failover peer "dhcp_opt6" {
secondary;
address 172.23.136.131;
port 520;
peer address 172.23.136.132;
- peer port 519;
+ peer port 520;
....

Node2(primary):
.....
failover peer "dhcp_opt6" {
- secondary;
+ primary;
+ mclt 1800;
+ split 128;
address 172.23.136.132;
port 520;
peer address 172.23.136.131;
- peer port 519;
+ peer port 520;
....

This fixed the issue on our systems.


Files

dhcpd-node1.conf (908 Bytes) dhcpd-node1.conf node1 secondary dhcp config Christopher Harbort, 09/30/2015 07:49 AM
dhcpd-node2.conf (908 Bytes) dhcpd-node2.conf node2 primary dhcp config Christopher Harbort, 09/30/2015 07:49 AM
Actions

Also available in: Atom PDF