| Subject: | isc-dhcp-4.2.3 - large memory usage of dhcpd in failover environment |
| Date: | Wed, 23 Nov 2016 02:43:46 +0000 |
| To: | dhcp-suggest@isc.org |
| From: | "Zhaolong" <qq470551101@126.com> |
Bug Report from www.isc.org:
Name: Zhaolong
Email: qq470551101@126.com
Software Version: isc-dhcp-4.2.3
OS: CentOS release 6.4 (Final) X86_64
Subject:large memory usage of dhcpd in failover environment
Bug Detail
===========
Hi there,
I'm using dhcp-4.2.3 with two groups of failover in my environment.
Usually dhcpd uses 1.2G on one host and 1.8G on the other.
But sometimes when I startup dhcpd on host 1, I found that the memory usage of dhcpd rises to 3.2G, and never falls.
But this does not happen everytime.
Each failover group contains 300,000-400,000 addresses.
My questions are:
1) Is there something wrong with my configuration? How to correct?
2) Is the memory usage of dhcpd normal? Why it is not stable?
3) Is there some tips about the memory usage of dhcpd?
I configured two ips on one netcard, like this:
host 1:
eth1: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc pfifo_fast state UP qlen 1000
link/ether 00:0b:ab:b6:ab:bf brd ff:ff:ff:ff:ff:ff
inet 10.250.2.110/24 brd 10.250.2.255 scope global eth1
inet 10.250.2.111/24 brd 10.250.2.255 scope global secondary eth1:1
inet6 fe80::20b:abff:feb6:abbf/64 scope link
valid_lft forever preferred_lft forever
host 2:
eth1: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc pfifo_fast state UP qlen 1000
link/ether 00:0b:ab:b5:d0:df brd ff:ff:ff:ff:ff:ff
inet 10.251.0.10/24 brd 10.251.0.255 scope global eth1
inet 10.251.0.11/24 brd 10.251.0.255 scope global secondary eth1:1
inet6 fe80::20b:abff:feb5:d0df/64 scope link
valid_lft forever preferred_lft forever
My failover declaration in dhcpd.conf:
(Notice that the split equals 255)
host 1:
failover peer "failover_changan" {
primary;
address 10.250.2.110;
port 1510;
peer address 10.251.0.10;
peer port 1510;
max-response-delay 60;
max-unacked-updates 10;
mclt 3600;
split 255;
load balance max seconds 3;
}
failover peer "failover_yanta" {
secondary;
address 10.250.2.111;
port 1511;
peer address 10.251.0.11;
peer port 1511;
max-response-delay 60;
max-unacked-updates 10;
load balance max seconds 3;
}
host 2:
failover peer "failover_changan" {
secondary;
address 10.251.0.10;
port 1510;
peer address 10.250.2.110;
peer port 1510;
max-response-delay 60;
max-unacked-updates 10;
load balance max seconds 3;
}
failover peer "failover_yanta" {
primary;
address 10.251.0.11;
port 1511;
peer address 10.250.2.111;
peer port 1511;
max-response-delay 60;
max-unacked-updates 10;
mclt 3600;
split 255;
load balance max seconds 3;
}
---
This email was received through isc.org Bug Submission Form