Even though the public cloud is a proven and well-tested model for numerous businesses, including ones we’re all familiar with like Google, Netflix and Amazon, there are still concerns about performance stability. That’s why it’s important to understand how cloud application providers fortify their services, and what, exactly, happens when the internet does go down.
To understand how companies should evaluate cloud vendor uptime and reliability, I sat down with Holly Wheeler, Director of Cloud Ops Service Availability at Genesys. With more than 20 years’ experience managing operations teams—supporting multiple technologies and enterprise accounts—she knows a thing or two about reliability.
Q: Some cloud solutions providers claim 99.999% uptime. How do I know if this is true?
Wheeler: It’s easy for a vendor to say they provide high availability but can they show it? A transparent cloud vendor will have a public-facing trust site that posts monthly platform availability percentages and all platform-wide incidents. Evaluate the status page on their trust site to see what information is published, and to see whether outages affect the entire platform or if they’re partial outages.
Q: What level of support should I expect from a cloud vendor?
Wheeler: First and foremost, a cloud vendor should value commitment to customer experience and transparency. Reliability and availability are not limited to platform performance. They include the vendor’s ability to fulfill commitments on time and at the agreed-upon price, as it pertains to all aspects of the customer journey. With respect to deployment and onboarding, you might require different levels of support, depending on the level of in-house resources you have. Check the available tiers of services and support options and understand if there’s flexibility to tailor those to your needs.
Q: What are some cloud provider best practices in terms of meeting SLAs for service availability, response time and remediation?
Wheeler: Many cloud providers still use monolithic architectures and a traditional operations model that requires repair time commitments of hours—not minutes. In contrast, cloud solutions that are architected with microservices technology and a DevOps model can contain issues when incidents occur, often reducing the scope of impact. Ensure the developers who write the code are also responsible for maintaining it. This way, when issues occur, they can be routed immediately to the appropriate subject matter expert, reducing the isolation and triage time that could traverse multiple tiers before they reach the person who can repair them. SLAs should be simple to understand; a good rule of thumb is 99.99% availability with fewer than five minutes per month of downtime.
Q: What penalties does Genesys face if they fail to meet SLAs?
Wheeler: As a leader in transparency, we publish our SLAs and credit policies on help.mypurecloud.com. If our service uptime falls below our guaranteed thresholds, our customers receive a credit on the following month’s invoice. We issue a 10% credit if uptime dips below 99.99%, and a 30% credit if it dips below 99%.
Q: What measures does Genesys take to mitigate availability risk?
Wheeler: Our cloud-native Genesys® PureCloudTM platform is built using a microservices architecture that limits impact scope. It’s built to leverage auto-scaling and self-healing practices to resolve incidents prior to impact. The cloud service spans multiple availability zones for active-active failover. And our DevOps model is orchestrated to engage subject matter experts at the onset of an incident; a thorough problem-management review process is leveraged in a blameless format to expand knowledge sharing among the teams. This helps disseminate best practices.
Built for Reliability
The right cloud platform can actually reduce downtime. A recent Forrester study indicates that customers on the PureCloud platform reduced downtime by 90%—from 20 hours to just 2 hours per year. Customers cited that when downtime did occur, the speed of recovery resulted in minimal business impact. Overall, reduced platform downtime equated to an avoidance of $120,000 in productivity loss.
To learn about other quantified benefits you can achieve with an advanced cloud contact center solution, download the Total Economic Impact of Genesys PureCloud by Forrester Consulting.