Data Center Fires
The recent data center fire which Terremark handled so well has had us thinking of other notable data center fires of the past couple years, and how important it is to implement an effective fire-suppression system.
Looming large in the recent history of data center fires is the one in Seattle’s Fisher Plaza last summer. It was, as we probably all remember, a huge, horrible mess, and it was their second fire in the past couple years. This one was apparently caused by a blown transformer, and it knocked out sites ranging from real estate, to recipes, to Bing Travel. The whole scene was an embarrassing mess for the data center, which had boasted as data centers do of its reliability and imperviousness to natural disasters. I remember footage of their clients carrying servers out of the building by hand… not a good scene for a mission-critical facility. To relive the horror in full, check out Seattle-based Tech Flash’s full coverage.
Late last year there was also a less catastrophic, but still problematic, fire at a Boston data center, also seemingly related to transformer issues. Since it affected brokerage and asset management firms, rather than all kinds of widely used sites like the Seattle data center, it wasn’t as widely covered in the media, but for more, see Data Center Dynamics’ article.
2008 also had its share of fiery events, specifically one in Texas and one in Wisconsin. The Planet’s fire in Texas didn’t do any actual damage to servers but did take out the power, resulting in downtime. Camera Corner/Connecting Point in Wisconsin, however, had a much more damaging fire… according to Data Center Knowledge’s article it actually destroyed 75 servers, routers, and switches. It also showed that the data center was not at all prepared for an event like that, and that they didn’t have fire suppression systems in place – something very important to check on when you are deciding on a service provider.
All these horror stories are in a way made all the more appalling by how well Terremark dealt with their recent fire. They stayed online throughout the entire affair, and no servers were damaged – in fact, only one electrical-gear cabinet sustained any damage at all. That Terremark can keep things together like that raises the bar for everyone else. Their backup plan was perfect and their response was fast and thorough enough to get everything under control before customers were affected.
Any thoughts on the subject of data center fires and disaster recovery? Did I forget any notable fires? We’re thinking of getting a whitepaper out on the subject of data center fire protection. What would people like to see in that?
Comments are closed