Corosync and pacemaker high availability cluster servers

Corosync Introduction:

Corosync can use a simple configuration file to define the way and protocol of information delivery when passing information. It is an emerging software, launched in 2008, but in fact it is not a real sense of the new software, in 2002, there is a project Openais, it is too large, split into two subprojects, which can achieve HA heartbeat information transmission function is Corosync, its code 60% Left or right from Openais. Corosync can provide a complete ha function, but to achieve more and more complex functionality, you need to use Openais. Corosync is the future direction of development. In later new projects, the general use of Corosync, and Hb_gui can provide a good ha management functions, can achieve graphical management. In addition, related graphics have RHCs suite Luci+ricci.

Pacemaker Introduction:

Pacemaker is a cluster manager. It leverages the messaging and membership capabilities provided by the preferred cluster infrastructure (Openais or heartbeat) for fault detection and recovery by secondary nodes and systems, and for the high availability of the implementation of the Cluster service (also known as resources). It can do clusters of almost any size, and comes with a powerful dependency pattern that allows administrators to accurately express the relationships between cluster resources (including order and location). Almost any script that can be written can be used as part of managing a pacemaker cluster.

Test Purpose:

Implementing Corosync+pacemaker for high availability management of servers

Test diagram:

Test ideas:

The first step: Configure the IP address and host name and other related information, to maintain consistency between the nodes.

Step Two: Install Corosync related Packages

Step three: Edit the configuration Corosync configuration document and check the log information to see if there are any errors

Step Fourth: Add Management Resources WEBIP

Step Fifth: Add Web Management Resource Services

Step Sixth: Add the Resource Management Group Web and add the Webip and webserver resources to the group for bundled use.

Step seventh: When the node fails, the node resource is transferred by default to another node (no DC is used)

Step Eighth: Test