Giters

haminhcong / jenkins-master-ha

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

Setup jenkins master high availability cluster note with PaceMaker

Model

Two nodes:

  • 192.168.123.10
  • 192.168.123.11

Install steps

  • Install packages and enable pcsd
<!-- perform both on two nodes 192.168.123.10, 192.168.123.11 -->
// Install package

yum install -y pacemaker pcs psmisc policycoreutils-python

// Disable selinux
setenforce 0
sed -i.bak "s/SELINUX=enforcing/SELINUX=permissive/g" /etc/selinux/config

// Enable pcsd

systemctl start pcsd.service
systemctl enable pcsd.service

// change passwd of hacluster user - username: hacluster, password: hacluster

passwd hacluster

Create And Setup Custom OCF agent

References:

Steps:

  • Copy file generic-script to /usr/lib/ocf/resource.d/heartbeat/generic-script
  • Copy files jenkins-master.yml and jenkins-master.sh to folder /root/jenkins-master-ha
  • Start pcs cluster: pcs cluster start --all
  • Create jenkins resource
pcs resource create jenkins-master \
 ocf:heartbeat:generic-script \
 script="/root/jenkins-master-ha/jenkins-master.sh" \
 statedir="/dev/shm" alwaysrun="yes" startandstop="1" \
 startargs="--start --docker-compose /root/jenkins-master-ha/jenkins-master.yml --container-name jenkins-master --startup-timeout 360" \ 
 stopargs="--stop --docker-compose /root/jenkins-master-ha/jenkins-master.yml --container-name jenkins-master --startup-timeout 360" \
 args="--monitor --container-name jenkins-master" \
 op start interval=0s timeout=360s op monitor interval=60s --wait
  • Verify cluster status: pcs status
  • Verify jenkins-master resource pcs resource show jenkins-master
[root@node-1 heartbeat]# pcs status
Cluster name: jenkins_master_cluster

WARNINGS:
Corosync and pacemaker node names do not match (IPs used in setup?)

Stack: corosync
Current DC: node-2 (version 1.1.19-8.el7_6.4-c3c624ea3d) - partition with quorum
Last updated: Fri May 24 00:54:52 2019
Last change: Sun Mar 17 20:40:35 2019 by root via cibadmin on node-1

2 nodes configured
1 resource configured

Online: [ node-1 node-2 ]

Full list of resources:

 jenkins-master	(ocf::heartbeat:generic-script):	Started node-1

Daemon Status:
  corosync: active/disabled
  pacemaker: active/disabled
  pcsd: active/enabled

At this time, jenkins-master container will be created in extractly one node in cluster.

Design cluster HA script for jenkins master single

  • Script: jenkins-master.sh
  • Docker-compose file jenkins-master.yml
  • Start args - startargs arg in pcs command: --start --docker-compose /home/centos/jenkins_master.yml --container-name jenkins-master --startup-timeout 90
  • Stop args - stopargs arg in pcs command: --stop --docker-compose /home/centos/jenkins_master.yml --container-name jenkins-master
  • Monitor args - args arg in pcs command: --monitor --container-name jenkins-master
  • startandstop arg value: 1
  • alwaysrun arg value: 1

With this setup

  • Start command: jenkins-master.sh --start --docker-compose /home/centos/jenkins_master.yml --container-name jenkins-master --startup-timeout 360
  • Stop command: jenkins-master.sh --stop --docker-compose /home/centos/jenkins_master.yml --container-name jenkins-master
  • Monitor command: jenkins-master.sh --monitor --container-name jenkins-master

References

About

Languages

Language:Shell 100.0%