Cloud hosting providers often describe their high availability in terms of “nines”.
At The C2 Group, we prefer a more binary approach: a "one" and two “zeroes”.
For the last three quarters, C2’s managed services team has provided 100% uptime for its clients in the cloud. What 100% uptime means is that there was no unscheduled or unplanned downtime during that span.
It’s difficult to guarantee 100% uptime as everything will fail at some point in time. While we depend on technology to be stable and predictable, we have to remember it’s also not infallible. Failures are at the heart of why an organization may experience downtime. Hard drives will crash, networks will go down, and people will make mistakes.
So how does C2 work to provide 100% uptime? It’s not without a little bit of luck, a good amount of planning, and a whole lot of redundancy.
Redundancy and automation are the most important factors for ensuring that downtime is kept to a minimum. When a hosting or cloud managed services provider mentions redundancy, think of this as a backup (or many backups). When redundancy is properly implemented with automation, and in the event of an outage or a failure, this ensures if a system or service goes down, a duplicate system (another server or virtual machine) can step in to pick up the load.
A couple ways C2 provides redundancy are through services like load balancing and cross-region replication. Load balancing is a service that distributes network or application traffic across multiple servers in the event of an unexpected increase of concurrent users. When paired with automatic scaling, the system can handle a sudden spike in traffic, otherwise load balancing will distribute volume across all available servers. Cross-region replication, meanwhile, triggers an automatic copy or backup of data to another datacenter in the event the primary copy becomes inaccessible due to a failure or outage.
Automation goes hand in hand with redundancy, as we automate the recovery or load distribution to remove humans from the equation. People are prone to making mistakes, and removing the human factor will help to make your systems more reliable. If you have processes that frequently require human intervention, automating these tasks and processes will not only remove risk, it will free these resources for more important tasks. Some good candidates for automation include applying Windows or server updates, generating backups, and pushing code deployments, to name a few.
While we can’t guarantee that implementing these processes and practices will always achieve 100% uptime, we can guarantee that you will be making great strides toward a more reliable and available infrastructure.
Remember, when it comes to managing your information systems, failing to plan is planning to fail... and plan on failures to happen.
Know the Nines
Uptime guarantees are usually bound by the terms of a service-level agreement, or, an SLA. Before committing to an uptime guarantee, it’s important to know what all those “9s” really stand for.
Three 9s or 99.9% uptime equates to almost nine hours of unplanned downtime per year. That’s roughly equivalent to closing your business for one day each year. In comparison, and while it may seem like splitting hairs, adding two more 9s (99.999%) amounts to just five minutes of downtime per year.
C2’s managed services team provides enterprise hosting and cloud management. For more information, or to receive a quote for your cloud needs, you can get in contact here.