Availability

In its simplest definition, availability is a measure of the amount of time a system or component performs its specified function.

99.95% availability allows for only 4 hours of unscheduled system downtime per year .

The percentage availability for any given server, application, or system can be raised by both increasing the time between system failures and reducing the time to recover from them. In practice this involves identifying and removing points of failure in four key areas:

Application Failures

Reasons that applications can fail include problems with inherent application reliability, mal-ware – software introduced either intentionally or unintentionally to disrupt a service – or human error.

Storage Failures

The data storage system must be protected from disk failure or disk controller hardware failures. Protection against storage failures includes the provision of resilient hardware, including RAID (Redundant Arrays of Independent Disks) disk configurations and fail-over hardware.

Server Failures

A server may appear unavailable for a variety of reasons, including operating system problems and hardware failures.

A faulty network connection may also result in a server or service being unavailable to it’s users.

In order to provide continuous application availability, a business must be protected against failures of both the server hardware, the operating system and application software. One typical mechanism to provide this protection is server clustering which provides a fail over capability in the event of a system problem.

Exchange 2010 abstracts the mailbox database and provides a mechanism to make the database itself resilient without the need to make the storage or the server nodes highly available.

 Site Failures

If the entire server infrastructure is at a single site, the failure of the site due to man-made or natural disasters can result in total system failure. Mitigation against site level failures in the case of Exchange 2010 may involve multiple database replicas across Database Availability Groups member servers in different sites.

Cheddon Limited provide design and implementation services for highly available systems and can provide guidance on options for backup, on site data/server recovery and full site Disaster Recovery.

Click here to contact us about Availability Enhancements for your business.