High Availability concepts
High availability and disaster recovery aim to minimize the downtime of applications so that customers and agents experience little or no interruption.
- High availability means that an application can immediately react to a problem with little or no operation loss. The reaction is often an automated process that allows systems to recover quickly from failures.
- Disaster recovery strategy is a mechanism intended to take over the contact center operation in a remote site in the event of disasters such as floods, power interruptions, or other disasters that can completely shut down the contact center operation at the main site.
High availability and disaster recovery use different concepts:
- Redundancy: Duplication of modules to increase the reliability of the system.
- Fault tolerance: Characteristic of a module that allows the module to operate at a reduced level, rather than failing completely.
- Load balancing: Distribution of workloads across multiple modules.
This document presents high availability and disaster recovery solutions at two levels:
- Modular clustering: Having properly coordinated redundant components to assume the operation in case of non-recoverable failure.
- Individual components: Manage the failure of individual components automatically by re-executing the requested operations or re-launching the components.
The methods imply some trade-offs in terms of hardware equipment, and architecture complexity. Some solutions may have high impacts on software licensing costs.
|
Note Each contact center has its own particularities. For example, the recovering time of a total failover site depends on the complexity of the architecture, the modules to recover, network topology, application configurations, hardware characteristics, and other distinguishing marks. |
High availability cluster architecture
A high availability solution tries to minimize the system downtime in contact centers. The solution has two strategies:
- Automatically manage individual failures in components by re-executing the requested operation or re-launching its execution process
- Having properly coordinated redundant components to assume the operation in case of non-recoverable failure.
The architecture is based on the following groups:
- Core cluster: Cluster with central services that distribute the media interactions for the agents and other functionalities.
- Telephony cluster: Cluster of servers to support voice media.
- Routing cluster: Cluster with routing solution that provisions intelligent routing and voice prompting.
- Operational cluster: Infrastructure that allocates the applications that support the contact center operations, and where the agents and supervisors connect with the specific front-end interfaces.