Kubernetes performance benchmarking from Speedscale

Speedscale achieves expertise as part of Google’s Partner Advantage Program, helping SaaS companies optimize cloud price-performance ratio while scaling quality Speedscale, the Kubernetes traffic replay company that can simulate production conditions, announces it has successfully achieved expertise in cloud-native application development as part of Google’s Partner Advantage Program. Achieving this expertise means that Speedscale has demonstrated strong customer success with Google Cloud. Nylas, a name in bidirectional email sync, used Speedscale’s traffic replay capabilities to optimize its cloud price-performance ratio and continuously deliver quality releases at scale. Using Speedscale, Nylas proved: that Google's Tau VMs with GKE delivers over 40% better price-performance than their current cloud architecture.

Email synchronization, the highest throughput part of their SaaS offering, improved 30x as part of their migration and continuous testing.

Testing at the speed of cloud-native

"Nylas needed modern observability and QA tools to deliver highly performant, quality releases, at the speed of cloud-native, which Speedscale allows us to do with confidence. With every new feature and every delivery, our developers add new nodes and microservices, more API integrations, and more interdependencies that, without proper testing, could easily trigger a breakdown in API communications. Fortunately for Nylas and our customers, Speedscale provides fast API performance benchmarking and testing needed to satisfy customer demand," said David Ting, Nylas’ senior VP of engineering, product, and CISO.

Speedscale demonstrates Kubernetes performance benchmarking with expertise in cloud-native app development with Google Cloud

Unit, UI, and manual testing are the industry standard. Yet according to the 2019 State of DevOps Report by DORA, the types of incidents that bring down production systems are often "caused by interactions between components." The vast majority of quality automation still requires an author to painstakingly create each individual test case and maintain them, which slows delivery dramatically. By leveraging traffic to auto-generate tests and environments, Speedscale allows users to exercise their APIs on GKE and finally keep pace with development. A user can also multiply traffic to simulate high load to test Google’s Autopilot and ensure auto-scaling performs as expected.

"Speedscale’s superpower is mocking real-world traffic patterns, including spikes in the volume of incoming requests and any variability in data returning from downstream services. It’s like testing in production with none of the risk. You simulate production conditions in pre-prod but also understand latency, throughput, and headroom of changes are within your expected bounds," said Ken Ahrens, co-founder of Speedscale.