High Availability & Fault Tolerance
A high availability server strategy should be an imperative for any organization with enterprise that contains or transacts mission-critical data. The term "high availability," however, can constitute many things and encompass a hydra of different techniques and technologies. So, what exactly is high availability, anyway?
High availability is the moniker applied to solutions that automatically failover to a designated backup system to allow an enterprise to continue work unabated. High availability provides an instant recovery point, meaning that recovery time is nearly immediate (less than 30 seconds). High availability as such is nearly synonymous with business continuity, and a complimentary strategy to disaster recovery. A company without a cohesive plan will fact immediate lost revenues, a tarnished reputation, and loss of business altogether.
There exist many components to an effective High Availability strategy. Paramount among these is adherence of best practices for change management, network operations, and internal and external services. Irrespective of the due diligence applied, networks will invariably fail. High Availability is about minimizing the impact when they do.
Three of the linchpins in a High Availability environment are Redundancy, Replication, Clustering, and Load Balancing. Implementing each of these on your network and web servers will provide your organization effective safeguards against potential downtime. Note that each of these can be implemented in either a physical or virtual manner. Which is most appropriate depends on the size and functional characteristics of a network.
- Redundancy is the policy of implementing a secondary system or network device that duplicates a server, and have it assume live operations if the main server fails.
- Replication refers to copying and distributing data and database objects from one database to another, and then synchronizing the data across both platforms. Like redundancy, its purpose is to provide an automated and seamless recovery point for data in the event of a failure.
- Clustering is a server technology that provides failover points at the hardware level. Servers configured in the same cluster provide transparent failover and operational continuity in case of failure.
- Load Balancing refers to splitting network traffic between servers to achieve best enterprise throughput and performance. Load balancing typically also is used with redundant servers to avoid a single point of failure.
Neteam's High Availability Implementation staff is fully accredited for Microsoft high-availability solutions. In our experience, due to their critical importance to businesses, high availability is most frequently required for sites employing Exchange server and SQL Server. Overviews follow of the native high availability options available for each of these cornerstone platforms.
Exchange Server features scores of powerful native high availability options. These include support for Single Copy Cluster, the traditional clustering model, Cluster Continuous Replication, a clustering technique that does not require shared storage. New for Exchange 2010 are Data Availability Groups (DAGs), which is comprised of discreet mailbox servers that, when designated as a member of a DAG, can copy databases to other mailboxes. DAGs support automatic database-level recovery from failures occurring on the database, server or network.
SQL Server supports several types of replications, which can transfer data to any ODBC-compliant data sources.
- Snapshot replication: transfers all table data from one database to another, every time the distribution agent runs.
- Transactional replication transfers only the changes detected to the subscribers and synchronizes the publisher and subscriber with minimal latency.
- Merge replication is a bidirectional replication mechanism that includes an internal conflict resolution agent to assure consistency across both data sources.
Neteam's staff of certified engineers is expert at devising and implementing rock-solid high availability solutions, regardless of the characteristics of your network. Our experience, and core knowledge of high availability assures your organization of the highest level of system and data safeguards, the smallest possible backup window, and the most attractive return on investment.
If your business relies on data, your business needs high availability. And Neteam knows high availability.