Verizon, Microsoft and Google were just some cloud providers to see their services interrupted so far this year from a variety of issues, from a change in the authentication system to a deadly winter storm. In the cloud computing era, some experts say we can only expect more outages — but with less severity.
Miles Ward, chief technology officer at Los Angeles-based Google partner SADA Systems, told CRN that cloud outages can prove less disastrous than when data centers have issues. With cloud-related issues, providers can fix the problem in parallel with a user’s team, whereas data centers can require an internal team to fix problems.
“Outages can mean the end for companies, depending on their choices in design and deployment, or they can be complete non-events,” Ward said. “Cloud has changed the nature of outages.”
As cloud adoption and the number of regions, zones and cloud services grow, everyone should prepare for more outages, Ward said. But he expects the type of global, all-service outages that garner headlines to decrease.
“Every cloud engineering team has seen how impossible it is for customers to engineer around these kinds of outages and is working hard to distribute, subdivide, and make fault-tolerant these central services,” Ward said. “The result may be a shift of focus where you might see even more minor failures in singleton services, while the global services survive seemingly unaffected by minor failures because of this investment in resilience.”
Companies today need copies of their data in distant regions, to run instances in multiple zones and automation to cut down on the time it takes to fix an outage, Ward said. At SADA, even demos are designed with high availability to run across Google Cloud and AWS.
In the meantime, CRN has collected a list of some of the largest cloud outages and issues to hit computers this year. Here’s what you need to know.
For more of the biggest startups, products and news stories of 2021 so far, click here.