By Tom Condon, RPA, FMA
From the October 2013 issue of Today’s Facility Manager
The seemingly endless range of disasters in the news lately, from Hurricane Sandy in 2012 to the recent flooding in Colorado, makes facility managers (fms) all over the country cringe at the thought of something like that happening to their facilities. Disaster recovery planning is something all fms must live with, but there is one aspect that many forget: technology. We all know that, in a disaster, we must be prepared to repair damaged facilities, relocate occupants, and, in extreme cases, find an entirely new location from which to operate. But the same disaster that damages a facility probably also impacts the facility’s technology.
While it is relatively straightforward to rent a space or hire a contractor to repair a facility, recovering technology from disasters can be a bit more complicated. Fms shouldn’t take it for granted that the IT staff has planned for disaster recovery the way an fm would, and here are some of the aspects of disaster recovery that all fms should be aware of.
The Cloud: With all of the attention cloud computing is getting today, it is easy to forget that cloud based technology is also a mainstay of disaster recovery planning. Cloud based systems can be particularly helpful for many fms, because they do not require an extensive technical knowledge of IT systems. There are two main ways to use the cloud for disaster recovery—buying software in a Software as a Service (SaaS) model and using it as an online backup tool.
Using the SaaS approach involves paying a monthly per user fee. The software vendor hosts the software, and users access it through the Internet with a web browser. Today, just about any kind of facility management software is available in the SaaS model, including CMMS, CAFM/IWMS, and even visitor management and access control.
The second way fms might use cloud technology as a disaster recovery tool is for online backup of a facility’s local data. Backup services will sell a client space on their data centers, and software installed on the client’s server will automatically backup data to the cloud as often as desired.
When local disasters pummel a facility and its data center, cloud systems will not be affected because they are not located on-site at the facility. Also, these systems are usually hosted in data centers designed for high availability and survivability. However, not all cloud systems are hosted in ideal environments. When signing up for SaaS, fms should make sure to ask about the provider’s data center uptime performance, where it is located, and whether it has multi-site redundancy in case a disaster affects the primary data center.
Local Backups: If fms choose not to use the cloud, then they need to back up data on some kind of media such as an external hard drive, USB drive, or tapes. Yes, believe it or not, tapes are still very common in this digital age, and these are surprisingly inexpensive and reliable.
No matter what kind of media is used for backup, fms should take care to not fall prey to these common mistakes: If using tapes, change the tape cartridge on a regular schedule, and rotate among a number of tapes so the workload is spread out evenly among them. Backup tapes do not last forever, so fms should check the manufacturer’s recommendations on when to replace them, and then do so. And no matter what media is used for backup, do not leave it sitting next to the server… that is a recipe for disaster. Many times in the case of a fire or flood, the backup media is destroyed along with the server. Place the media in a fireproof and waterproof safe, or better yet, take them off-site completely.
There is a variety of software available that helps fms plan for disasters and how to deal with them. One product on the market features planning templates based on various industries. This software helps users to create a plan and then publish and communicate it. Meanwhile, other systems focus on managing emergencies, such as a product that provides automatic alerts based on weather warnings and other threats. These types of systems may also incorporate mass notification capabilities that deliver messages to individuals by continuing to call, e-mail, and/or text them until they respond. There are mobile versions so fms can manage the incident from smartphones or tablets out in the field.
Another way to communicate quickly is with social media. As noted in my July 2012 column, some fms are using Twitter to communicate. One fm sent a tweet to alert occupants that a tornado had been spotted nearby. This message got through even though the power in the facility was out (and the phone system was useless).
And what happens if you need to replace technology infrastructure completely after a disaster? If the system is not cloud based, and you need to get a ruined data center up and running quickly, one option is to lease a data center housed in a shipping container. The HP EcoPOD, for example, is a self-contained unit with cooling that can provide the IT capacity of a large data center in a very small footprint. These can be located in areas not suitable for traditional data centers, like warehouse space or even parking lots (see my September 2012 column for new approaches to data centers).