Building Multi-Tenancy ML Applications with GKE and Istio (Cloud Next '18)

Опубликовано 26 июля 2018, 0:16

Microservices and container orchestration, such as Kubernetes, have enabled a new wave of scalable applications to be built. However, they create interesting challenges when isolating multi-tenant workloads. In this session, you’ll learn how Descartes Labs built their multi-tenant GCP native platform on GKE, leveraging the Istio service mesh. We’ll cover approaches to solving challenges such as service-to-service authorization, telemetry, rate limiting, and how to isolate each customer’s code while still allowing them to scale to thousands of CPU’s. Finally, we will demo some examples that our own machine learning team has built over Petabytes of satellite imagery.

IO263

Event schedule → g.co/next18

Watch more Infrastructure & Operations sessions here → bit.ly/2uEykpQ
Next ‘18 All Sessions playlist → bit.ly/Allsessions

Subscribe to the Google Cloud channel! → bit.ly/NextSub event:


re_ty: Publish; product: Cloud - Containers - Google Kubernetes Engine (GKE); fullname: Tim Kelton, Sam Skillman; event: Google Cloud Next 2018;