09-20-13 | Blog Post
As you watch with amazement and awe of the devastation seen with Colorado’s massive flooding, you can’t help but realize that you can never predict a natural disaster. Unfortunately, this might only be the start as climatologists predict an increase in coastal flooding and 100-year, 500-year storms.
In Some data center operators take their chances with floods, Arielle Emmet states how IT leaders are handling these dire predictions. There are a variety of expert opinions on what data center managers should do to protect themselves from potential disasters. Some experts suggest that data centers need to harden their existing facilities. Others advise moving data centers to higher ground or outside of the floodplain. While other experts say to follow both suggestions.
One thing that all can agree on is to – think long term.
Data Center Lessons
So what can Michigan data centers learn from flooding disasters like Hurricane Sandy and the Colorado flooding?
Consider Geographic Locations – Many times organizations are more focused on an area’s real estate costs or the cost of power when deciding on a data center location. However, these organizations need to stop and also think about the region’s probability of natural disasters.
There seems to be a strong concentration of data centers on the West and East Coasts as well as in the Southern Coastal Regions. And these regions suffer from many risks like hurricanes, tornadoes, flooding, landslides, earthquakes, etc.
Data centers located in Michigan have a great advantage over facilities in these regions. Michigan offers some of the lowest risks in hurricanes, earthquakes and tornadoes. We also offer cooler temperatures, which can lower data center cooling costs.
Strategic Data Center Design – If you are located in a flood-prone area, strategic data center design is extremely important. You need to consider the location of critical equipment including generators and generator fuel pumps. We saw in the case of Hurricane Sandy, energy blackouts happened as a result of data centers that could not fuel their generators because flooding ruined their fuel pumps that were located in the basement.
High Availability Infrastructure – Using high availability architecture will reduce your risk of losing Internet connectivity or power. Your primary power circuit should be provided by the primary UPS and be backed up the primary generator. A secondary circuit should be provided by the secondary UPS, which is backed up by the secondary generator. This redundant design ensures that a UPS or generator failure will never interrupt power in your environment.
If you are using a managed service provider, you will want to ensure that all hardware is configured for high availability, including dual power supplies and dual NIC cards.
Redundant Design – A fully redundant data center design allows automatic failover and ensure server uptime in the event of a natural disaster. This includes using multiple ISPs and fully redundant networks with automatic failover. Pooled UPS, battery and generators will ensure a backup source of power in the event one provider fails.
Use Cloud Computing – With virtualization, your virtual server can be copied or backed up to an offsite data center, and spun up on a virtual host in minutes in the event of a disaster. Since the virtual server is hardware independent, the operating system, applications, patches and data can be safely transferred from one data center to a second site without reloading each component of the server.
For more information on geographical natural disaster risks, read Michigan Data Center Earthquake Risks, Michigan Data Centers Offer Little Hurricane Risks, Low Tornado Risks for Michigan Data Centers, and The Right Distance for Michigan Disaster Recovery Sites.