Highly Available System Architecture Concepts

iSCSI SAN Storage Software / Features / StarWind’s Active-Active HA storage architecture / Highly Available System Architecture Concepts

Highly Available System Architecture Concepts

What is high availability and what steps must be taken when designing system architecture to support it? Given the tools and technologies available in today’s information technology environments, it is possible to design and build a solution that offers no downtime, assuming you take great care to design reliability and availability into all aspects of the system. StarWind software architecture is designed to avoid any known single points of failure, but it is also important to plan for high availability failover across the entire system architecture. This begins at the most elementary level with the server hardware itself by making sure that network traffic into your system can be automatically routed to other available servers in the cluster, should one go down for whatever reason.

From there, the system architecture should focus on data security, which can be accomplished through a variety of high availability storage design solutions. A variety of RAID configurations can be used to ensure data reliability and recoverability across multiple storage devices and software technologies.


High availability network design is an important facet of any highly available system architecture. Considerations such as cooling via air conditioning on the floor, redundant power supply within the facility as well as in/out of the facility (or highly reliable reserve power supplies or generators), and even protection against fire and water damage are all design issues handled by the largest and best hosting facilities today.

Besides planning for network, software, hardware, and storage failures, the best high availability architectures also provide solutions for application restarts and data recovery in the event of a complete system or partial subsystem failure that may have left part of the system in an invalid state or left data in an incomplete or invalid state. The possible way to avoid failures such as these is to provide software and/or hardware solutions whose goal is to detect and understand the invalid state of a system while also providing the facilities for bringing the system and/or data back to its correct state so that proper operations can continue.

In other words, architecting a high availability system or high availability failover environment is not just about keeping your environment up…the best architectures will also plan for what must happen after a failure occurs.