04 Jun 2012

5 tips for making your systems reliable

On the list of priorities for any IT system, reliability takes a prominent place.Whatever else systems do or don’t achieve, those that break down, fail, or otherwise under perform are guaranteed to interrupt your business – usually at the worst possible time.If you’re looking to make your systems more reliable, here are a few simple tips that can help.UpdateEven if your systems are functioning well, be sure that you have rigorous processes in place to carry out regular updates.In addition to their obvious benefits for security, these will save you from trouble down the track while making sure that your systems are more supportable.MonitorProactive monitoring is one of the most powerful tools in the reliability arsenal.From storage to memory consumption, watching for early warning signs and acting well before your systems are likely to encounter trouble is a must. Watch your logs and feeds carefully and take action to make them error free.Introduce redundancyMost failures can be avoided through careful redundancy planning – a task that virtualisation has made easier. From network links to data storage, running multiple systems in parallel will greatly reduce the chance of an outage.RespondOf course, it’s difficult to make a system fail proof, and unexpected failures will happen.When they do, limit the damage by making sure that your IT team or provider is following best practice procedures.The ITIL standard, for example, draws a careful distinction between the ‘incident’ (that is, the business interruption) and the ‘problem’ (the unknown or undiagnosed cause of the incident). The approach it suggests is to focus on the business outage, and not the IT one.The first step in an ITIL-aware solution is to get the business back on line. If printing has stopped, then the immediate imperative for the Incident Manager might be to get a new printer and restore business functionality. The point is to restore business function as quickly as possible, so that the business is not kept waiting for the underlying problem to be identified and solved.Make it someone else’s problemOf course in the age of the cloud, one of the best ways to get reliable systems is to opt for cloud-based services with strong Service Level Agreements (SLAs).Alternatively, the enormous array of managed services on offer will also often be more reliable replacements for those managed in house. Stephen Sims is a Brennan IT’s General Manager – Sales and Marketing.