Data Center Disasters: How to Prepare for the Worst
Data centers safely house essential computing equipment and important information. Their server racks can contain many millions of dollars’ worth of equipment and value in the data that they contain. Companies go to great lengths to safely back-up data and prevent/fix any hardware failures to minimize downtime. Still, one of the most overlooked threats to customers and companies is being unprepared for data center disasters.
When a company is building a new data center, disaster avoidance and recovery is a major subject of concern. Large data centers are often in locations where there is a low risk of devastating natural events. Even so, data center managers need to know how to prepare for the worst, and what to do after disaster strikes.
Types of Disasters
The first step in any disaster avoidance and recovery plan is identifying the potential types of disasters that can occur. The following are all examples of major disasters that a facility must prepare for.
- Fire – A fire is one of the most common types of disasters a data center can face. Facilities have to prepare for the risk of a wildfire approaching them by protecting their property and of course, prepare for a fire starting from the inside.
- Flood – Floods can be caused by excessive rain, melting snow, broken damns, or a number of other natural events. Another type of flood to be aware of comes from broken pipes within the data center.
- Earthquake – Earthquakes can be absolutely devastating, and often hit with very little warning. Even small earthquakes can cause serious damage to unprotected equipment.
- Tornado – A tornado’s powerful wind can knock out power, cut data circuits, push trees over onto a building, and much more.
- Hurricane – Everyone knows that computing devices don’t fare well with water, but on top of that, hurricanes can cause wind damage, fires, power outages, and much more.
- Terrorist Attack – Data centers are responsible for supporting a large portion of the economy, which makes them a potential target for cyber attacks.
Data center managers will need to evaluate the likelihood of each of these disasters. Understanding that earthquakes, for example, are much more likely in areas built on and around fault lines can help in understanding where to invest disaster recovery resources.
Types of Impacts a Disaster Can Have on Data Centers
When a disaster strikes it can have an immediate and long-lasting impact on a data center. Preparing for the risk of a disaster is important, but risks that follow the event deserve the same focus.
For example, if there’s a fire, the obvious impact is that equipment could burn up and become unusable. In an extreme fire that is out of control, the entire facility could be destroyed. A good data center, however, will respond quickly to even a small fire in order to minimize further damage. If the data center doesn’t have an electronics-safe fire suppression system, the damage from the water could be extensive.
Every type of disaster will have ‘primary’ impacts (such as the fire itself) and then ‘secondary’ impacts that need to be planned for. When done properly, a data center can minimize the damage, downtime, and impact a disaster has to the facility.
How to Prepare for a Disaster
Preparing for a disaster starts from the very beginning of any data center design. Whether building a new facility from the ground up or simply converting a current space, this will take some thought. The following are key disaster planning solutions that can help to dramatically reduce the risk to any data center:
- Proper Fire Suppression – Fire suppression systems are required in almost all buildings. Choosing a system that won’t damage the computer equipment within the data center is important. There are many solutions on the market today to protect servers from fires such as a fluoroketone fire protection fluid.
- Advanced Physical Security – Physical security will help to minimize the risk of terrorist attacks, help to prevent a disgruntled employee from causing damage, and improve overall safety.
- Seismic Server Rack Shelves – In areas where earthquakes are possible, specially designed server rack shelves help to keep the servers, routers, switches, and other equipment in place. They will also help to reduce the vibration that reaches this equipment, which can minimize the risk of damage.
- Flood Management System – Data centers often have raised floors that will help to keep water from reaching the equipment. This system will also include pumps that can quickly remove water and drain it outside the facility.
- Multiple Data Circuits – Multiple data circuits in a facility will provide redundancy in the event that one of them is cut or damaged due to a disaster.
- Redundant Power Source – Having an uninterruptable power supply system in place will keep a data center running even when commercial power is out for an extended period of time.
- Disaster Recovery Location – In situations where downtime is absolutely unacceptable, a company can operate multiple data centers in geographically distant locations.
Reacting to a Disaster in a Data Center
When a disaster hits, all data centers should have a well-rehearsed disaster recovery plan in place that’s easy to follow. This plan will guide employees on what to do during the disaster, as well as the days and hours after. A good recovery plan has instructions for evaluating damaged systems, how to restore all essential systems, and much more.
Employees should also know what systems to restore first, and which ones can wait. In most cases, restoring internet connectivity will be a priority because it is integral for for all other system functions.
Companies that place a priority on up-time and reliability will have a detailed disaster recovery plan that they update regularly. They will also perform mock disaster events to ensure everyone knows how to respond to situations that they hope will never occur.