Chapter 1. High Availability Add-On overview
The High Availability Add-On is a clustered system that provides reliability, scalability, and availability to critical production services.
A cluster is two or more computers (called nodes or members) that work together to perform a task. Clusters can be used to provide highly available services or resources. The redundancy of multiple machines is used to guard against failures of many types.
High availability clusters provide highly available services by eliminating single points of failure and by failing over services from one cluster node to another in case a node becomes inoperative. Typically, services in a high availability cluster read and write data (by means of read-write mounted file systems). Therefore, a high availability cluster must maintain data integrity as one cluster node takes over control of a service from another cluster node. Node failures in a high availability cluster are not visible from clients outside the cluster. (High availability clusters are sometimes referred to as failover clusters.) The High Availability Add-On provides high availability clustering through its high availability service management component,
1.1. High Availability Add-On components
The High Availability Add-On consists of the following major components:
- Cluster infrastructure — Provides fundamental functions for nodes to work together as a cluster: configuration file management, membership management, lock management, and fencing.
- High availability service management — Provides failover of services from one cluster node to another in case a node becomes inoperative.
- Cluster administration tools — Configuration and management tools for setting up, configuring, and managing the High Availability Add-On. The tools are for use with the cluster infrastructure components, the high availability and service management components, and storage.
You can supplement the High Availability Add-On with the following components:
- Red Hat GFS2 (Global File System 2) — Part of the Resilient Storage Add-On, this provides a cluster file system for use with the High Availability Add-On. GFS2 allows multiple nodes to share storage at a block level as if the storage were connected locally to each cluster node. GFS2 cluster file system requires a cluster infrastructure.
LVM Locking Daemon (
lvmlockd) — Part of the Resilient Storage Add-On, this provides volume management of cluster storage.
lvmlockdsupport also requires cluster infrastructure.
- Load Balancer Add-On — Routing software that provides high availability load balancing and failover in layer 4 (TCP) and layer 7 (HTTP, HTTPS) services. The Load Balancer Add-On runs in a cluster of redundant virtual routers that uses load algorithms to distribute client requests to real servers, collectively acting as a virtual server. It is not necessary to use the Load Balancer Add-On in conjunction with Pacemaker.
1.2. Pacemaker overview
Pacemaker is a cluster resource manager. It achieves maximum availability for your cluster services and resources by making use of the cluster infrastructure’s messaging and membership capabilities to deter and recover from node and resource-level failure.
1.2.1. Pacemaker architecture components
A cluster configured with Pacemaker comprises separate component daemons that monitor cluster membership, scripts that manage the services, and resource management subsystems that monitor the disparate resources.
The following components form the Pacemaker architecture:
- Cluster Information Base (CIB)
- The Pacemaker information daemon, which uses XML internally to distribute and synchronize current configuration and status information from the Designated Coordinator (DC) — a node assigned by Pacemaker to store and distribute cluster state and actions by means of the CIB — to all other cluster nodes.
- Cluster Resource Management Daemon (CRMd)
Pacemaker cluster resource actions are routed through this daemon. Resources managed by CRMd can be queried by client systems, moved, instantiated, and changed when needed.
Each cluster node also includes a local resource manager daemon (LRMd) that acts as an interface between CRMd and resources. LRMd passes commands from CRMd to agents, such as starting and stopping and relaying status information.
- Shoot the Other Node in the Head (STONITH)
- STONITH is the Pacemaker fencing implementation. It acts as a cluster resource in Pacemaker that processes fence requests, forcefully shutting down nodes and removing them from the cluster to ensure data integrity. STONITH is configured in the CIB and can be monitored as a normal cluster resource. For a general overview of fencing, see Section 1.3, “Fencing overview”.
corosyncis the component - and a daemon of the same name - that serves the core membership and member-communication needs for high availability clusters. It is required for the High Availability Add-On to function.
In addition to those membership and messaging functions,
- Manages quorum rules and determination.
- Provides messaging capabilities for applications that coordinate or operate across multiple members of the cluster and thus must communicate stateful or other information between instances.
kronosnetlibrary as its network transport to provide multiple redundant links and automatic failover.
1.2.2. Configuration and management tools
The High Availability Add-On features two configuration tools for cluster deployment, monitoring, and management.
pcscommand line interface controls and configures Pacemaker and the
corosyncheartbeat daemon. A command-line based program,
pcscan perform the following cluster management tasks:
- Create and configure a Pacemaker/Corosync cluster
- Modify configuration of the cluster while it is running
- Remotely configure both Pacemaker and Corosync as well as start, stop, and display status information of the cluster
- A graphical user interface to create and configure Pacemaker/Corosync clusters.
1.2.3. The cluster and pacemaker configuration files
The configuration files for the Red Hat High Availability Add-On are
corosync.conf file provides the cluster parameters used by
corosync, the cluster manager that Pacemaker is built on. In general, you should not edit the
corosync.conf directly but, instead, use the
cib.xml file is an XML file that represents both the cluster’s configuration and the current state of all resources in the cluster. This file is used by Pacemaker’s Cluster Information Base (CIB). The contents of the CIB are automatically kept in sync across the entire cluster. Do not edit the
cib.xml file directly; use the
pcsd interface instead.
1.3. Fencing overview
If communication with a single node in the cluster fails, then other nodes in the cluster must be able to restrict or release access to resources that the failed cluster node may have access to. This cannot be accomplished by contacting the cluster node itself as the cluster node may not be responsive. Instead, you must provide an external method, which is called fencing with a fence agent. A fence device is an external device that can be used by the cluster to restrict access to shared resources by an errant node, or to issue a hard reboot on the cluster node.
Without a fence device configured you do not have a way to know that the resources previously used by the disconnected cluster node have been released, and this could prevent the services from running on any of the other cluster nodes. Conversely, the system may assume erroneously that the cluster node has released its resources and this can lead to data corruption and data loss. Without a fence device configured data integrity cannot be guaranteed and the cluster configuration will be unsupported.
When the fencing is in progress no other cluster operation is allowed to run. Normal operation of the cluster cannot resume until fencing has completed or the cluster node rejoins the cluster after the cluster node has been rebooted.
For more information about fencing, see Fencing in a Red Hat High Availability Cluster.
1.4. Quorum overview
In order to maintain cluster integrity and availability, cluster systems use a concept known as quorum to prevent data corruption and loss. A cluster has quorum when more than half of the cluster nodes are online. To mitigate the chance of data corruption due to failure, Pacemaker by default stops all resources if the cluster does not have quorum.
Quorum is established using a voting system. When a cluster node does not function as it should or loses communication with the rest of the cluster, the majority working nodes can vote to isolate and, if needed, fence the node for servicing.
For example, in a 6-node cluster, quorum is established when at least 4 cluster nodes are functioning. If the majority of nodes go offline or become unavailable, the cluster no longer has quorum and Pacemaker stops clustered services.
The quorum features in Pacemaker prevent what is also known as split-brain, a phenomenon where the cluster is separated from communication but each part continues working as separate clusters, potentially writing to the same data and possibly causing corruption or loss. For more information on what it means to be in a split-brain state, and on quorum concepts in general, see Exploring Concepts of RHEL High Availability Clusters - Quorum.
A Red Hat Enterprise Linux High Availability Add-On cluster uses the
votequorum service, in conjunction with fencing, to avoid split brain situations. A number of votes is assigned to each system in the cluster, and cluster operations are allowed to proceed only when a majority of votes is present.
1.5. Resource overview
A cluster resource is an instance of program, data, or application to be managed by the cluster service. These resources are abstracted by agents that provide a standard interface for managing the resource in a cluster environment.
To ensure that resources remain healthy, you can add a monitoring operation to a resource’s definition. If you do not specify a monitoring operation for a resource, one is added by default.
You can determine the behavior of a resource in a cluster by configuring constraints. You can configure the following categories of constraints:
- location constraints — A location constraint determines which nodes a resource can run on.
- ordering constraints — An ordering constraint determines the order in which the resources run.
- colocation constraints — A colocation constraint determines where resources will be placed relative to other resources.
One of the most common elements of a cluster is a set of resources that need to be located together, start sequentially, and stop in the reverse order. To simplify this configuration, Pacemaker supports the concept of groups.