A Message from Duane Barnes, SVP, Technology (@duanebarnes)
My name is Duane Barnes, SVP, Technology (@duanebarnes) here at RapidScale. Frequently we are brought into Disaster Recovery opportunities by our sales partners around the country, as DRaaS is often looked at as a gateway service to the cloud. This tends to be the case because customers can purchase DR services and implement them with very little to no impact to their production environments in most situations. I liken the DR sales process to selling insurance. For many companies, it is simply seen as an expense that rarely pays any return. With that said, when I get a chance to speak with customers directly, I will ask a fair number of questions about their production environments. I like to start there because more often than not their production environment is not in a good state to begin with. Where is it hosted today? Are there single points of failure? Do you have hardware and software maintenance contracts? What does the network connectivity to the environment look like?
The whole concept of DR planning is about mitigating risk. What is more likely to happen: hardware failure on a single server sitting in an office building somewhere causing an outage and/or data loss, or a natural disaster taking out an entire Tier III data center? The former is definitely the answer. Now, that is not to say buying DR services is a waste of money, or having the dialogue is a waste of time. For the majority of customers, it is an eye-opening experience when they have the epiphany that they really shouldn’t be housing a server in a broom closet in their office where the AC shuts off at 5 p.m. on Fridays and turns back on at 8 a.m. on Mondays.
I fully realize the timing of this blog post is during the aftermath of a massive hurricane that was pounding the Texas coast. It is certainly not my intention to come off as an ambulance chaser. For that reason, I’m focused more on the topic of production rather than DR. Consider this example: ABC Customer has all their servers hosted internally and a natural disaster occurs. Sure, with DRaaS they can fail their services over to RapidScale and within 4 hours we will have them back online. However, if they had migrated those servers to the cloud to begin with, their data would be more resilient and accessible from anywhere already. One of the biggest challenges during a real disaster is the logistics of employees accessing data remotely. If servers are already in the cloud, and access to them is already part of their day-to-day processes, then that is one less thing to worry about in a true disaster scenario. This is one of the main reasons why Desktop as a Service is a very good option to consider in production and DR situations. Many of our customers that purchase CloudRecovery from us also purchase CloudDesktop – the reason being the scenario mentioned above. We can scale out desktops in the cloud on demand during an event to enable customers’ employees that may be displaced during a disaster to access all of their applications and operate as if there were sitting in their office.
In closing I’ll restate the obvious. Mitigate your highest risk items first. If you have servers sitting in your office, perhaps DR should come second to migrating those servers to the cloud first where you do achieve a return on the expense. If you’re already in the cloud or simply don’t have time for a migration due to a pending event (such as Hurricane Harvey), then absolutely look at a DRaaS solution.