The Hostway Blog

Choosing the Right Data Center

With matters of security, support, accessibility and redundancy in mind, it’s crucial to choose the right data center for your company. Whether you’re building out your own infrastructure, colocating or hosting with a provider, your data will ideally reside in this location for years. Not asking the right questions could result in a failure that could cost your business a significant amount of time and money.

In the past year, we’ve seen major data center outages from big names like Amazon, Dropbox, Microsoft and Google. The 49 minutes of downtime for Amazon cost the company an estimated $5 million in missed revenue.

For an in-depth look at what choosing the wrong data center could cost you, review Ponemon’s 2013 Study on Data Center Outages. When speaking with data center managers, 84% said they would rather walk barefoot over hot coals than experience data center downtime, but 91% of them actually experienced an unplanned outage in the past 24 months.

We all know that technology fails, but it’s important to choose the most reliable data center that meets your needs. But how do you make the choice? According to InformationWeek Editor Kurt Marko, “CIOs have a bad case of the paradox of choice: Faced with myriad alternatives – private data centers, leased wholesale space, colocated facilities, managed infrastructure services, virtual private clouds, public IaaS, PaaS, SaaS – they get overwhelmed and become paralyzed.”

So what should you look for when choosing the right data center?

  • Security: In the case of hosted data centers, security tops the list of critically important capabilities to consider. On the human side, the facility should offer passcode entry and 24x7 video surveillance inside and outside of the building. The data center you choose should be equipped with the necessary certifications for your business. In addition, the expert staff should be able to ward off cyber attacks which are continuously on the rise.  And if you are subject to any regulatory restrictions, you want to ensure the data center is capable of meeting your compliance requirements.
  • 24x7 Technical Support: This includes certified, highly experienced engineers and managed server monitoring to alert your staff and the hosting provider of any potential server issues. The provider should also offer a troubleshooting and remediation service to identify root causes of any problems for quick resolution. Response and resolution times are paramount.
  • Accessibility:  Remote access is crucial, especially in the case of a crash. Does the data center offer “low-level” server access which is still functional even if the operating system is offline? When it comes to the staff at the data center, only those with clearance credentials should be allowed to access the more secure parts of the facility. In addition to accessibility for your server, you want to be certain that the data center staff will be accessible to you. Direct to data center phone numbers that allow you to speak to the technicians who are located with your servers is an added bonus.
  • Cooling and Power: Similar to natural disaster security, the environment should have plenty of space to handle the servers it is hosting, but also be able to support and protect them. Ask about the number of backup generators available, and how their emergency systems are set up to support a maximum capacity of servers. UPS battery failure (55 percent) and UPS capacity exceeded (46 percent) were two of the most common root causes of data center outages in 2013.
  • Hardware: Choose a data center that utilizes the latest technology and operates with reliable brands that are not prone to equipment failure. You choose dependable brands when making business-related purchases, so make sure that the data center you select does as well.
  • Network: Data centers need to have fast networks that reduce latency experienced by the end user as much as is possible. Furthermore, network engineers should have plans in place to ensure that when failure occurs, the data center’s Mean Time to Recovery is as quick as possible.
  • SLAs: Don’t be shocked by the fine print. When choosing a data center, do your due diligence and read the service-level agreement (SLA) to make sure that the terms agreed to are fair to both parties.
  • Geographical Location: Protection against natural disasters such as hurricanes and floods is a must. For example, Facebook built a data center near the Arctic Circle to take advantage of a climate that offers natural cooling resources. In addition, consider multiple data centers that are spread out amongst multiple locations, like New York and San Francisco, in order to afford even more redundancy.

Finally, make sure that you develop a trusted advisor relationship with your provider or the owner of your data center. Technology fails; what’s important is that it is limited and responded to quickly. Ensuring that your data center team feels that their success is intertwined with yours will help build a stronger relationship and lead to continued IT success for your company.

SOURCES
http://www.infoworld.com/slideshow/107783/the-worst-cloud-outages-of-2013-so-far-221831#slide3
http://www.emersonnetworkpower.com/en-US/Brands/Liebert/Infographics/Pages/Causes_of_Downtime.aspx