Network Deployment (Distributed operating systems), v8.0 > Establishing high availability
High availability manager
The product includes a high availability manager component. The services that the high availability manager provides are only available to product components.
A high availability manager provides several features that allow other product components to make themselves highly available. A high availability manager provides:
- A framework that allows singleton services to make themselves highly available. Examples of singleton services that use this framework include the transaction managers for cluster members, and the default messaging provider, also known as the service integration bus.
- A mechanism that allows servers to easily exchange state data. This mechanism is commonly referred to as the bulletin board.
- A specialized framework for high speed and reliable messaging between processes. This framework is used by the data replication service when the product is configured for memory-to-memory replication.
A high availability manager instance runs on every application server, proxy server, node agent and dmgr in a cell. A cell can be divided into multiple high availability domains known as core groups. Each high availability manager instance establishes network connectivity with all other high availability manager instances in the same core group, using a specialized, dedicated, and configurable transport channel. The transport channel provides mechanisms which allow the high availability manager instance to detect when other members of the core group start, stop, or fail.
Within a core group, high availability manager instances are elected to coordinate high availability activities. An instance that is elected is known as a core group coordinator. The coordinator is highly available, such that if a process that is serving as a coordinator stops or fails, another instance is elected to assume the coordinator role, without loss of continuity.
Highly available components
A highly available component is a component for which a high availability group is defined on the processes where that component can run. The coordinator tracks high availability group membership, and knows on which processes each highly available component can run.
The coordinator also associates a high availability policy with each high availability group. A high availability policy is a set of directives that aid the coordinator in managing highly available components. For example, a directive might specify that a component runs on a specific process, if that process is available. Directives are configurable, which makes it possible for you to tailor policies to your installation.
The coordinator is notified as core group processes start, stop or fail and knows which processes are available at any given time. The coordinator uses this information, in conjunction with the high availability group and policy information, to ensure that the component keeps functioning. The coordinator uses the policy directives to determine on which process it starts and runs each component. If the chosen process fails, the coordinator restarts the component on another eligible process. This reduces the recovery time, automates failover, and eliminates the need to start a replacement process.
State data exchange
The high availability manager provides a specialized messaging mechanism that enables processes to exchange information about their current state. Each process sends or posts information related to its current state, and can register to be notified when the state of the other processes changes. The workload management (WLM) component uses this mechanism to build and maintain routing table information. Routing tables built and maintained using this mechanism are highly available.
Replication
The data replication service (DRS) provided with the product, is used to replicate HTTP session data, stateful EJB sessions, and dynamic cache information among cluster members. When DRS is configured for memory-to-memory replication, the transport channels defined for the high availability managers are used to pass this data among the cluster members.
Related
When to use a high availability manager
Core groups (high availability domains)
High availability groups
High availability group policies
Core group View Synchrony Protocol
Set up a high availability environment
Transactional high availability