High Availability Series: General Concepts

By January 17, 2013High Availability

In the world of IT, “high availability” is a term you often encounter. There are a few things that go into having high availability and assessing it. This week we will cover the general concepts behind maintaining highly available environments.

There are two major things to consider with high availability: redundancy and separation.

Redundancy

Redundancy involves providing excess capacity in the design in order to account for any failures without a performance decline.

An example of redundancy would be taking  a server and plugging it into not just one, but two power circuits to protect against the power failure of one.

high availability: two circuits power source

But what if the server itself fails? Add another server and put them together in a cluster.

high availability: server cluster on two circuits

However, if the circuit fails, the server or cluster would still go down. The key to protecting your availability is to double up (or triple even) on your equipment and power sources.

Separation

Diversifying the power sources will help protect your servers from going down due to malfunction or power outages.  A good way to improve your server’s or server cluster’s availability is to connect it to power sources from two different circuits. To maintain availability even during a widespread power outage, we recommend using at least one uninterruptible power supply (UPS). These UPSs are powerful for protecting availability because they take very little time to assume the power burden if there has been a mains power failure. A UPS is able to quickly supply energy through batteries or a flywheel.

high availability: server cluster on two circuits with UPS

While redundancy and separation are two different elements to ensuring high availability, it’s important to note that you need both. A server failing is as likely as a power failure. A solid plan for high availability accounts for both redundancy and separation in order to ensure there is a plan B for any situation.

Mirazon has experienced and skilled technicians to help engineer your data centers and networks for high availability. Contact us today to find out how we can help you prevent downtime due to power outages or hardware malfunctions.