A Closer Look at Cloud Backup and Cloud Outage

A Closer Look at Cloud Backup and Cloud Outage

January 15

Amazon Web Services (AWS) provides service to a lot of different large companies like Pinterest, Instagram, and Netflix. On Christmas Eve, there was an issue with one of the AWS data centers that shut down Netflix for thousands and effectively ruined Christmas for all of them.

Ok, maybe Christmas wasn’t ruined, but Christmas Eve is one of the most popular nights for watching movies and the fact still stands that the cloud broke down and that shouldn’t happen. Ever.

Whether you run a business from a cloud platform, use it for cloud storage, or rely on online software, you should never have to worry about something going wrong in the cloud. The problem is that every time you take advantage of the cloud in one form or another, you must put faith in the company you’re using and your destiny in their hands. It’s worth it to partner with someone you know has your best interests in mind from the start.

Let’s look at an example. Suppose you’re using StorageCraft for cloud backups, you shouldn’t have to worry about an interruption during the transfer process, or worse yet, what if you’re running a virtual machine from the StorageCraft cloud while you attempt to get your physical production units online? You already suffered one interruption so you shouldn’t have to worry about another problem occurring on our end. We take the utmost care in insuring that unplanned outages do not occur. After all, we’re experts in the business continuity er… business.

Last fall I wrote an article about the various capabilities featured in StorageCraft Cloud Services to illustrate what sort of safeties were in place to ensure secure and uninterruptible access to your cloud backups. As I read through it, I realized that a few of things just sounded like fancy marketing words. Realistically, how protected is the data, what do these words mean, and what’s really keeping disruption from occurring?

When you take advantage of cloud backup with StorageCraft Cloud Services[i], your data is safe and accessible 24/7 with high availability N+1 redundant infrastructure for power, cooling, and internet access.

Well, that sounds dandy, but what does N+1 mean?

N+1 is a form of resilience that ensures system availability in the event of primary component failure. It’s basically a simple math equation where N represents the components and +1 represents at least one backup component. N+1= at least one level of redundancy. This means that not only do we have high-quality primary components, we also have a backup system for power, cooling, and internet access.

What about a site destroying event like a fire? Couldn’t disasters like these feasibly destroy primary and secondary components?

Theoretically, yes. But, to protect from fire, our data centers are equipped with FM200 Gaseous Fire Fighting Systems. Ok, what the heck is that? FM200 is essentially the gold-standard in fire-fighting systems. It’s designed to safely extinguish electronic equipment cleanly and quickly. FM 200 is stored as a liquefied gas and is non-conductive so it won’t damage electrical equipment. But, since the data center’s load, security, humidity, and temperature are monitored 24/7, it’s highly unlikely that fire would ever be an issue to begin with.

Awesome, what about hurricanes and earthquakes?

The data center locations were selected for their geological stability, as well as the historical absence of natural disasters in the area. These areas are located in the eastern and mid-west regions of the United States and are considered disaster-safe. You can also choose to store a portion of your extra-critical data in both data centers, just in case. StorageCraft Cloud Services gives you the option to mirror your data in both data centers if you wish, regardless of what tier of services you’ve selected.

Ok, what about user error?

As we in the backup and disaster recovery business know, user error tends to be one of the biggest causes of problems. In fact, the latest Amazon Web Services outage on Christmas Eve was caused by user error. Essentially, an engineer ran a process that inadvertently deleted necessary information thereby knocking out parts of the system, and thousands of Netflix viewers.

What can we say about user error? It happens because nobody is perfect, but being in the business we have intimate knowledge of how a simple deleted file can cause a major disaster. That means we take paramount care in determining the best people to take care of the data centers, and we make sure they keep things closely monitored around the clock.



[i] This article focuses on the data centers in the United States.