The term High Availability has become popular with regard to information systems over the last few years. Simply put, High Availability (or HA) refers to systems that are designed to provide services that are accessible by their users without interruption.
Identifying Single Points of Failure and eliminating these failures by providing redundancies can be a challenge when designing an HA system. To account for maximum availability and uptime, it is necessary to automate the transfer of services from a failed resource to a redundant resource.
A Single Point of Failure can exist at many levels in a complex system; ranging from hardware failure, loss of network from a service provider or a software failure on a server. Redundancies must be addressed at each level to ensure High Availability. Hardware redundancies are a common best practice today and are already implemented by many organizations. However, many of these same organizations have not yet implemented solutions for Application Level availability.
Designing a High Available Environment
High Availability starts with a solid foundation: the infrastructure. Multiple redundancies for power, cooling and network connectivity need to be in place. A simple way to accomplish this is to host the environment in a datacenter that already has these provisions. Server hardware also needs multiple redundancies. To mitigate hardware as a single point of failure, a virtualization solution (VMware, for example) can be implemented. This will distribute the logical servers across several host servers and leverage a redundant Storage Area Network (SAN.)
Even with a robust and highly available infrastructure, downtime can occur for software and services within the environment. A well designed and healthy server still requires regular maintenance for security patching, updates to functionality and software upgrades. Since downtime and the possibility of software failure will always exist, it is necessary to implement High Availability at the Application Level for critical systems.
Exchange High Availability
Most organizations view email as a critical application. Fortunately, Microsoft Exchange 2010 has robust High Availability features built in. This technology is known as a Database Availability Group (DAG.) The DAG allows each mailbox database to operate independently of an individual server by replicating changes in near real time to a database copy on another server. Exchange’s Active Manager Technology monitors each server and determines, based on performance counters, the most desirable location for the active database. In the case of a server failure, Exchange leverages Microsoft’s Failover Clustering technology to ensure that the database copies are activated on the redundant server. Creating a single namespace for client access using a Client Access Server Array Object (CAS Array) will further improve availability. To protect against a complete site loss, a server can be placed in a geographically divergent datacenter and the DAG can be extended to that site.
High Availability requires careful planning and redundancies at many levels. As a trusted IT partner, DP Solutions can assist with planning, architecting and implementing a robust High Availability Solution for your organization. This includes a complete discovery of existing technologies, business processes and interdependence of services in order to design the correct solution. Implementation includes testing of systems to ensure a robust and highly available environment. Additionally, failover scenarios are documented to ensure minimal disruption to business continuity in the case of systems or infrastructure failure. To learn more about our High Availability solutions, contact your DP Solutions’ representative at 800.679.4377 or firstname.lastname@example.org.