11 Data centre preparations for severe weather
The winter of 2014-2015 has seen record-setting snowfall across much of the U.S. As described in a New York Times article published in mid-February, “Boston’s Winter from Hell,” Boston has been especially hard hit. Since the Times article was published, Boston received even more snow, and as of March 4 the city was just four inches shy of its all-time snowfall record.
Boston's mass transit system, the T, has been largely or completely out of service for significant periods. Businesses "have been hammered," according to the Times, and many employees have faced significant challenges just getting to work.
Digital Realty has sites in several areas that received unusually heavy snowfall this winter, including Boston, New York, and Chicago.
Since our data centres are at the centre of many of our clients' IT strategies (and hence their strategies for growth), data centre downtime due to weather events is simply not an option. We understand the risks that severe weather poses to utility infrastructure and mission-critical operations, and we implement proactive emergency preparedness procedures and staffing levels to minimise these risks.
I recently sat down with Jim Palmieri, our Vice President of Technical Operations, and Danny Lane, our Vice President of Property Operations, to learn more. (Danny's operations team, which includes our data centre staff, is the largest team in the Digital Realty organisation; Technical Operations is the team of engineers that supports Danny's operations team, and provides technical support if any issues arise with the power or cooling in the data centre.)
Preparations to ensure uptime
Jim and Danny noted that their teams keep a constant eye on the weather forecast, and if they see extreme weather coming, the teams swing into action to implement a well-defined set of preparations to minimise the possibility of downtime:
- All affected sites – as well as any neighboring sites –are put on alert.
- The team implements a staffing strategy which ensures round-the-clock onsite coverage - including making onsite and offsite lodging arrangements in the event site access is at risk. In addition, both in-market and regional resources are identified in case the event duration should extend beyond initial expectations and additional staff rotation is required.
- Emergency contact lists are reviewed and distributed to site teams.
- The team makes sure each site's extensive vendor and supplier resource pool is at the ready - for example, the snowplow timetable is confirmed, and critical vendors - such as vendors related to the HVAC (heating and cooling), generator, and uninterruptible power supply (UPS) systems - are mobilized or stationed onsite to make sure these resources will be available.
- Regional and national vendor contact lists are distributed to team members.
Site & Equipment
- Emergency generator fuel tanks are topped off. This is perhaps the most important precautionary measure, because utility outages from downed power lines are a very real possibility.
- Backup power generators are tested.
- Additional site tours and rounds are conducted in order to ensure the integrity of all critical electrical and mechanical systems.
- The frequency of inspections of outside areas is increased.
- Snow is shoveled off roofs, to minimise the potential for leaks.
- The team confirms that adequate supplies of water, first-aid kits, flashlights, etc. are on hand.
Jim and Danny added that the above severe weather preparedness procedures are supported by on-going procedures such as conducting regular emergency drills and proactive cross-training of personnel so they can work at alternate locations.
In the second part of this series, we'll look at emergency preparedness from a different angle - the specific ways in which our clients benefit from working with a provider with a large data centre footprint.
Rebecca Bergman, Director of Corporate Communications (Rebecca_DLR)