How to prevent storage outages

No business likes to think of their systems going down, particularly if the issue originates in the data storage system, since data is the lifeblood of a business. This showcases why it’s essential to maintain your data storage systems carefully.

There are four key areas to consider when it comes to managing your data:

1. Capacity allocation

This refers to how well disks are provisioned to ensure your business has the capacity it needs, when it needs it, without overspending on unused storage. Ideally, it should be possible to get the storage you need ‘just in time’, but virtualising storage arrays may be the best approach to ensure you don’t unexpectedly run out of capacity.

2. Capacity utilisation

This refers to how well the storage is being managed so that useless data isn’t retained (and paid for), and to ensure that data is stored on the most appropriate type of media. For example, disk arrays can be an expensive way to store that data that you need to keep but rarely access. In this case, tape is a cheaper and more reliable way to store data. Archiving to tape and practising basic data hygiene, such as deleting duplicates and other data you don’t need, can help you reclaim expensive disk space.

3. Data protection

Given the importance of data, it’s essential to make sure it’s protected. Data needs three layers of defence, protecting against: corruption or loss due to errors or malicious activity; machine failure; and facility outages or disruptions. Virtualising your storage infrastructure lets you create layers of replication and mirroring at a reasonable cost and in a way that’s easy to manage.

4. Storage performance

Performance is important because it’s what users see, so most IT departments are looking to improve speeds. Storage hypervisors can help. However, poor performance isn’t necessarily due to the storage system itself; it’s often connected to the application code or the server hypervisor.

Despite your best efforts, systems may go down and you may end up needing to implement your disaster recovery (DR) plan. If this happens, it will help to know that you have a plan and that it has been well-tested.

A good DR plan should include six key elements:

1. Redundancy: Server virtualisation has removed the need for duplicate server hardware, making redundancy more cost effective. It is now easier to shift key servers between sides for additional capacity and maintain operations in the event that primary systems fail.

2. Business continuity: Many hosting providers offer highly-scalable platforms on which companies of all sizes can build DR environments. If a disaster occurs at the main business premises, key systems can be redirected to the hosted environment and the business can resume operations as quickly as possible.

3. Documentation: It’s important to methodically document your IT infrastructure including related support and technical contacts, and server and application configurations. Developing recovery time objectives (RTO) and recovery point objectives (RPOs) is also key.

4. Prioritised business functions: You should detail each IT function and the likely impact should it be interrupted. Mapping out internal dependencies and potential alternative suppliers or staff that can be brought on if necessary, can also help businesses function in the face of a disaster.

5. Clear internal communications: Employees need to know what to do to resume operations as quickly as possible. It’s equally important to communicate with suppliers and partners to ensure all parties have active DR plans in place that offer support to one another.

6. Testing and optimisation: Most importantly, when it comes to business continuity, it’s vital to test your business recovery plans in real-world scenarios and use the insights that emerge to continually optimise your DR plan.