When public cloud services go wrong

By Hank Smith
Posted in Virtualization
On May 20, 2013

By now we are all familiar with cloud-based services. Whether they be public, private, or hybrid offerings, cloud-based services are prevalent in many organizations. There are many cloud providers to choose from, and over the past two years many of Gotham’s customers have begun migrating services to the public cloud, typically starting with email.

Recently a Gotham customer migrated their email from an internal Exchange 2007 environment to the public cloud. The overall migration went smoothly; however problems started post-migration with network latency issues communicating with the cloud provider. As we all know, email is a core business service. Companies rely on email for communications, so this greatly impacted users’ productivity as they experienced connectivity issues to their hosted mailbox.

According to the cloud provider, all of their customers were affected, and they were instructed by Microsoft to upgrade to Exchange 2010 SP3, which had only been out for 2 weeks at the time. Gotham’s customer had no say in this, and following the upgrade to SP3, the issues only got worse. Upon asking the cloud provider what the rollback plan was, our customer discovered they didn’t have one. When asked about change management, the provider didn’t have anything for that either. It’s worth noting that this was not a startup cloud provider, but one that had been around for a number of years. The issue was finally resolved by adding more CAS servers to the cloud-based environment. However, intermittent mailbox access impacted user productivity at Gotham’s customer for approximately 3 weeks.

The moral of the story is, once your company goes with a cloud provider for a business-critical service like email, you are very much at the mercy of that provider to maintain the shared service. It’s important to do your homework and ask the right questions before moving any core service to the cloud.

Questions you should ask include the following:

  • How many years has the company been in business?
  • What is their change control procedure? Have the hosting company provide in writing what the change control is.
  • What is the overall uptime? Many hosting companies state 5 “9s”, i.e., 99.999% uptime. Have them clarify what exactly that means to them.
  • Ask if the hosting company supports multiple links to their site in the event that one of the links goes down.
  • Request customer references.
  • Ask about SLAs for support tickets.
  • Ask about the security of the solution; and regulatory compliance (e.g., HIPAA, Sarbanes-Oxley)
  • Ask about the hosting facility itself (e.g., physical security, redundant power, Internet)
  • How does the hosting company handle updates to the environment such as OS and Exchange updates?

As seen above, having been in business for several years doesn’t necessarily mean much. Seemingly correct answers to the other questions might not mean much either, so you should ask all of these questions and consider the totality of the responses before selecting a service provider.

Hank Smith

Hank Smith

Hank is an expert in the design, development, and delivery of cost-effective, high-performance technology solutions. As manager of Gotham’s Virtualization practice, Hank builds motivated, productive teams for Gotham’s large-scale networking and infrastructure engagements, develops implementation standards and methodologies around virtualization technologies, and manages overall design and implementation of multiple infrastructure projects.