Apache OpenWhisk + Kubernetes: A Perfect Match for Your Serverless Platform Ying Chun Guo guoyingc@cn.ibm.com Zhou Xing xingzhou@qiyi.com Agenda • What is serverless? • Kubernetes + Apache

with a Bang! Multiple Approaches Kubeadm Improve Project X Focus on Kubernetes only Design better abstractions Work with small number of teams at a time • Setting up the whole Go-Viet infrastructure Scalable • In the case of traffic spike, for instance, we can spin up new containers much more quickly than setting up new VMs. Faster MTTR Project X was used as deployment tool for Go-Jek International Uptime • System resources like CPU, memory, etc. are more effectively utilized in container world than in VMs. Efficiency • Automatic service discovery allows engineers to not maintain any configuration

Fast, insight-driven decision-making in highly dynamic and dangerous conditions is, more often than not, the difference be- tween life and death. Booz Allen’s innova- tive edge computing solution, SmartEdge heart rate, I can check yours too, look after my team better, adjust my tactics ac- cording to environmental and physical fac- tors. This drives better outcomes.” Historically, to confirm a team has received

topology (sites, affinity groups, NUMA, etc.). ​This session will explain the options to gain better performance, resource optimization and availability through tuning of vSphere, and Kubernetes configuration needed to meet requirements • Don’t assign an odd number of vCPUs • Never compose a VM larger than the number of physical cores A NUMA aware hypervisor can have IO benefits too For the vSphere

vSphere topology (sites, affinity groups, NUMA, etc.). This session will explain the options to gain better performance, resource optimization and availability through tuning of vSphere, and Kubernetes configuration is needed to meet requirements • Don’t assign an odd number of vCPUs • Never compose a VM larger than the number of physical cores A NUMA aware hypervisor can have IO benefits too For the vSphere

What did we do? 1. Write first Dockerfile 2. Reduce image size 3. Don’t run installs 4. Use better base image 5. Write .dockerignore 6. Improve incremental speed 7. Switch to use a Maven plugin What did we do? 1. Write first Dockerfile 2. Reduce image size 3. Don’t run installs 4. Use better base image 5. Write .dockerignore 6. Improve incremental speed 7. Switch to use a Maven plugin What did we do? 1. Write first Dockerfile 2. Reduce image size 3. Don’t run installs 4. Use better base image 5. Write .dockerignore 6. Improve incremental speed 7. Switch to use a Maven plugin

sidecar container as dependent environment • Encapsulate API / SDK of other tools using image for better integration/ collaboration • Leverage k8s integration capabilities, such as external service registration • Improve development experience, such as CLI, plugin for IDE, dev on Cloud • Move forward to better DevOps under micro-service architecture • Consolidate the consistent events and data, and create

provided by host kernel are not satisfied • OpenStack is too complex • Unified infrastructure • Better isolation VM related Projects Virtlet KubeVirt RancherVM Kata Container Focus : deploy REAL

Why not RollingUpdate ? Ø What are the advantages of batch gray release ? • more reliable and better control • More flexible • More efficient StatefulSetPlus StatefulSetPlus Service (Kube-proxy

conversion, API doc, encoding/decoding etc 5.gRPC based interface (e.g. CRI) 2.How we can do better to use Kubernetes? 1.Programming Patterns in Kubernetes 1.this is the main difference of Kubernetes