Heartbeat 3 Group Knowledge (1) Cluster principle

I. The concept of HA cluster

(1) nodes (node) run the heartbeat process of a stand-alone host, called nodes, nodes are the core components of HA, each node running the operating system and heartbeat software services, in the heartbeat cluster, the nodes have primary and secondary points, respectively called the Master node and standby/ Backup nodes, each with a unique hostname and a set of resources of its own, typically running one or more application services on the master node. And the standby node is generally in the monitoring state.

(2) A resource (Resource) resource is an entity that can be controlled by a node, and when a node fails, these resources can be taken over by other nodes, and in heartbeat, the entities that can be used as resources are: disk partitions, file systems, IP addresses, application services, NFS file systems

(3) event is a cluster of things that may happen, such as node system failure, network connectivity failure, NIC failure, application failure, and so on. These events can result in the transfer of resources from the node, and the HA test is based on these events.

(4) The Action (action) event occurs when HA response, action is controlled by the shell, for example, when a node failure, the backup node will be programmed by the execution of the script to shut down or start the service. And then take over the resources of the failed node.

Two, heartbeat composition

(1) Heartbeat structure heartbeat1.x and 2.0.x version of the structure is very simple, each module is concentrated in the heartbeat, to the 3.0 version, the entire Heartbeat project was split, divided into different projects to develop separately. heartbeat2.0.x Previous versions have modules:

Heartbeat: Communication detection module between nodes

HA-LOGD: Cluster Event Log service

CCM (Consensus Cluster Membership): Cluster member consistency management module

LRM (local Resource manager): Native Resource Management module

Stonith Daemon: Causes the problematic node to detach from the cluster environment

CRM (Cluster resource Management): Cluster resource management module

Cluster Policy engine: Cluster policy engine

Cluster transition Engine: Cluster transfer engine

(2) The components after Heartbeat3.0 split:

Heartbeat: The original message communication layer is independent of the Heartbeat project, the new Heartbeat is only responsible for maintaining the information of cluster nodes and their prior communication;

Cluster Glue: The equivalent of an intermediate layer, which is used to correlate heartbeat and pacemaker, consisting mainly of 2 parts, namely LRM and Stonith.

Resource Agent: A collection of scripts used to control service startup, monitor service status, which will be invoked by LRM to implement various resource launches, stops, monitors, and so on.

Pacemaker: That is, cluster Resource Manager (referred to as CRM), which manages the entire HA control center, the client configures the management to monitor the whole cluster through Pacemaker. Pacemaker provides a variety of user management interfaces, respectively, as follows: (1) CRM Shell: A character based management method;

(2) A web Konsole window using the AJAX Web configuration method;

(3) Hb_gui, that is, heartbeat GUI graphics Configuration tool, this is the original 2.1.x GUI configuration tool;

(4) DRBD-MC, a Java-based configuration management tool.