In June 2021, a massive cloud service outage caused many high-profile websites to go down. Learn how companies can be better prepared for such outages.
Many visitors to Amazon, PayPal, Reddit, and other popular websites were greeted with a “Service Unavailable” error message in early June 2021. A massive cyberattack didn’t take down these sites but rather a cloud service outage.
The sites were using Fastly’s content delivery network (CDN), an Infrastructure as a Service (IaaS) solution. CDNs are used to improve the reliability and performance of high-traffic websites. On June 8, however, a previously unknown software bug caused Fastly’s CDN to fail. This cloud service outage resulted in Amazon, PayPal, and many other high-profile sites to go offline.
You don’t have to look hard to find other examples of cloud service outages. Just nine days after the failure at Fastly, an outage in Akamai Technologies’ CDN caused the websites of E-Trade, Southwest Airlines, United Airlines, Westpac Bank, and many other companies to go down. A routing table issue was to blame.
Cloud service outages are not restricted to IaaS solutions. Other types of cloud services can experience them, too. For instance, Salesforce — a provider of Software as a Service (SaaS) solutions — had cloud service outages in both May 2021 and June 2021 due to Domain Name System (DNS) issues. As a result, 150,000+ businesses worldwide were unable to use Salesforce’s customer relationship management (CRM) services.
Although cloud service outages often impact large numbers of businesses, they are typically short-lived. For example, Fastly’s outage lasted only 49 minutes, whereas Akamai Technologies’ and Salesforce’s services were offline for several hours.
Dealing with the Disruption
As the Fastly, Akamai Technologies, and Salesforce outages demonstrate, cloud services are not immune to outages. Companies concerned about the downtime and other hardships caused by a cloud service outage might be hesitant about keeping their business operations in the cloud. This is a valid concern. Even short outages can be disruptive — and there’s little companies can do to prevent them. However, companies can take measures to lessen the impact if one occurs.
For starters, businesses should develop a contingency plan for each cloud service they use. In it, they should address how to continue operations if that cloud service goes down. The specific actions will depend on the type of cloud service being used and a company’s resources. Answering the following questions can help determine the actions to take:
- Do employees have what they need locally to perform essential job duties if the cloud service goes down? For instance, do they have local copies of apps, files, and contact lists?
- Does the cloud service provider offer the option of storing apps, data, or other resources in multiple availability regions or zones? Storing resources in multiple regions or zones means that employees can continue to work if there is a cloud service outage in one of those locations.
- Is it feasible to have a secondary cloud service provider if the primary provider experiences an outage? Having a secondary provider might be a viable option for some cloud services. For example, many SaaS providers offer free versions of their online software, so setting up secondary accounts would be a cost-effective way of ensuring continuity for services such as email and video conferencing. In other cases, though, having a secondary provider would be cost prohibitive (e.g., having two CDN providers).
- Does it make sense to use a hybrid cloud deployment? Hybrid cloud deployments contain both private and public clouds. The private and public clouds operate independently, but apps or data move between them.
After companies create their contingency plans, they must make sure all the elements needed to carry out those plans are in place. For example, if a business decides to have secondary email service provider, that cloud service needs to be set up so it is immediately available if the primary one goes down. Plus, the company needs to let employees know about the secondary service and provide training on how to use it.
Be Better Prepared
Cloud service outages are not uncommon, so it is important to have contingency plans in place to lessen their impact. We can help you develop and implement customized plans for the cloud services used in your business.
Software Bugs flickr photo by FastJack shared under a Creative Commons (BY) license
