5 tips on outsourcing your IT infrastructure
On Friday 27 March, two provinces in the Netherlands were hit by one of the largest power outages in Dutch history, leaving a million households without electricity. Schiphol airport saw flight cancellations, trains stopped running, people were trapped in elevators, and business premises were evacuated. The cost, which ran into millions, was caused by a defect in a high-voltage substation.
This power outage made it perfectly clear just how vulnerable we are and how much we rely on electricity. At many companies, the staff was unable to work and faced major disruption. They were overcome by a feeling of helplessness, especially when it became apparent that the cell towers had also failed. Many organizations immediately wondered about the state of their business continuity, because without power, how can you keep your IT structure and applications up and running? Processes grind to a halt, loss of business and damage to reputations are inevitable, and there are considerable costs associated with starting up again. The solution is simple: outsource your IT infrastructure to a data center and shift from on-premises to off-premises. But what should you look for in such a case?
- Tier 3 data center
This is the basis. Data centers are classified into tiers 1 to 4. The higher the tier, the better the availability of the data center. Most companies find that tier 3 data centers offer the best price-quality ratio. Thanks to redundant components and various independent distribution paths for power and cooling, these data centers have a maximum annual downtime of 1.6 hours. By way of comparison, a tier 2 data center is allowed up to a maximum of 22 hours of downtime per year. If you require even more security, choose a provider with multiple data centers that are geographically dispersed, which will enable you to set up a twin data center solution to further reduce the risks. - Redundant IT equipment
With only a tier 3 data center, you’re not quite there yet. The chance of downtime is reduced further by implementing a redundant hosting solution like separate power feeds, multiple hard disks in a RAID setup, and double network switches and uplinks. If your company operates on an international level, do not underestimate the effects of a global network. Using redundant network infrastructure and connectivity, part of the network setup can temporarily be out of order, for example as a result of a power outage, without any serious consequences. - Proven security measures
How reliable is a hosting provider? Look at the certifications that they have. The right certificates prove that their processes are in order and that they are regularly checked. This is the only way of knowing that there are established procedures in place to prevent, intercept, and resolve problems. - 24-hour support
You can’t entirely rule out problems, as the recent power outage showed. It is therefore a good thing if engineers are present day and night to tackle issues immediately as well as to ensure that communication is prompt and clear, for example through a Network Operations Center (NOC), through email, and through social media. - Service Level Agreements
A Service Level Agreement (SLA) contains the qualitative service levels that you can expect from a hosting provider. Check the SLA properly beforehand: is availability guaranteed when it really matters? How quickly does the service provider need to respond? Who is liable if something goes wrong? Also, regularly read through the SLA so that it is kept up to date.