The Cloud, and what is oftentimes denoted within the Information Technology Industry as Software as a Service (SaaS), has developed to the point where many small to medium sized businesses have come to rely heavily on its infrastructure. SaaS offers solutions previously only available to large businesses to the SMB market by offsetting cost through centralized cloud-based hosting. Amazon Web Services (AWS), among the most popular of these offerings, provides hosting even to larger enterprises, like Netflix, Reddit, Business Insider, and Slack. Although Cloud infrastructure like that offered by Amazon, Google, and Microsoft is incredibly robust, no system has 100% uptime. This became very apparent during the recent large scale AWS outage.
The root cause of the outage was incredibly banal—a mistyped command meant to move some cloud hosted servers in one of Amazon’s datacenters was mistyped. That typo took down or severely impaired the operation of some very large companies, including those listed above: Netflix, Reddit, Business Insider, and Slack—to name a few. A large number of SMBs were affected as well. The outage was localized, effecting only servers and applications hosted in Amazon’s East Coast datacenter. Normally, large enterprises will spread their infrastructure across multiple datacenters, even if they’re hosted in cloud infrastructure like AWS. That this outage was able to cripple a company as large as Netflix—a company valued as of this writing at 32.9 billion dollars—shows a glaring flaw in the organizations business practices. For SMBs, the trade-off is cost.
Spreading your infrastructure across multiple environments may seem like an unnecessary expense, especially considering a service with uptime figures of 99%, an outage that falls within that 1% would account for 88 hours.
How much money would your business lose if it was unable to function for three and a half days?
With any technology, it’s important to have a backup. NASA and other aerospace industries use the term “fail-safe.” This means that if any one system goes down, it’s capable of safely failing over to a backup system. The savvy SMB owner of course understands the importance of a secure backup solution, to protect their business from data loss. What happens, however, when the cloud hosted application where all of that data is stored isn’t available for several hours, or several days? Even the best backup isn’t worth much if the software needed to utilize the data in the backup is unavailable.
Before considering migrating to a wholly cloud hosted solution, it’s important to consider the ramifications of down time.
If you’re hosting directly with a service like Google or Amazon the additional expense of spreading critical data and applications across multiple datacenters can be worth far more than simple peace of mind. In other cases, it may make more sense to consider a hybrid or private cloud solution, allowing you, the SMB owner, to have far more control over where your data is stored, what backups are in place for that data, and what happens in the event of an outage.
No technology offering is a panacea. As cloud offerings like Amazon Web Hosting, Google Cloud, and Microsoft Azure have centralized more and more data into fewer and fewer physical and virtual locations, they have also introduced more and more single points of failure. Before jumping into the Cloud or SaaS with both feet, consider if you might be allowing a single mislaid keystroke to cripple your business.