Businesses are increasingly dependent upon electrical power to operate. It’s common sense that you should guard your IT systems against the likely event that they are rendered powerless. You only need to take a look at the live list of power outages to see that there are a number of outages happening at any one time and that you should prepare accordingly.
The problem with a power outages isn’t necessarily the instant impact it has on your building and IT systems, because many power outages are short-lived. The bigger impact is, in fact, noticed when power is restored. Hardware (especially old hardware) doesn’t respond well to power surges or interruptions. Switches and routers commonly fail due to fluctuations in voltage and current. When power is restored, your servers may all come back up in good working order, but without your switches and routers they are rendered useless and productivity can’t continue until these have been replaced.
So how do you protect yourself from the possible downtime caused by your next power interruption?
- Outsourced IT
If your IT systems are outsourced through a MSP (managed services provider) then your IT systems will be held in a data centre somewhere. You should check on the power resilience built in by the data centre to ensure you are happy with this. We still regularly come across instances where IT systems have been taken down by power failures at data centres, that even their methods of resilience can’t protect against. According to some research by Emerson Network Power ,91% of data centres experienced unplanned outages in the past 12 month.If you are worried about losing power then your best option is to have a copy of your IT systems held independently of your MSP so you can quickly failover to your standby system should your data centre have a power outage. Backups won’t help in this instance as you’ll have no IT systems to run them on. You should, however, always have a backup copy of your data held somewhere independently so in the worst case scenario, at least you won’t lose your data. IT systems can be replaced (in time), data can’t.
If you manage everything in-house, then you have the option of building in some resilience against power surges and short term outages by means of UPS’s. But bearing in mind that these can still fail, and during prolonged outages these will not maintain your IT systems for long enough, you should have another plan. Again, a secondary standby set of systems that can run in a separate geographical location, which won’t be affected by the same power cut, would be advantageous. You should also maintain spare hardware (routers, switches, HDD’s) ready for when you come to failback to your live systems after power is restored. The cost of standby systems and spare hardware though can start to escalate and this is where an independent cloud provider can start to reduce costs thanks to hardware syndication. The time and resource options of managing Disaster Recovery should also be considered, as managing and maintaining a second set of systems will incur a resource cost.
In summary, power failures are difficult to protect against unless you have a Disaster Recovery solution in place that can provide instant failover. Consider what you would do if your production environment were without power for a few hours. Are you happy to wait it out and deal with the consequences afterwards? It all boils down to whether you want to take the risk that a power outage won’t hit you.