cncf overview and building edge computing using kubernetes
4 likes429 views
The document discusses the concept of cloud-native computing using Kubernetes and its application in edge computing. It highlights the benefits of microservices, scalability, resource optimization, and the need for real-time data processing closer to data sources for critical applications. Various technologies and platforms for edge computing are mentioned, emphasizing the shift from centralized to distributed models.
cncf overview and building edge computing using kubernetes
- 1. CNCF & Building Edge Cloud using Kubernetes Presenting - We are Kumar Brothers⊠Krishna Kumar & Sanil Kumar
- 3. 3 evolveâŠ!
- 4. 4
- 5. 5
- 6. 6 Containers Cloud Native Cloud Native: CNCF (2015) âąCloud native computing uses an open source software stack to: â segment applications into microservices, â package each part into its own container â and dynamically orchestrate those containers to optimize resource utilization Open Source IaaS PaaS Open Source PaaS Virtualiza- tion 2000 2001 2006 2009 2010 2011 Non- Virtualized Hardware 2013 2015 IaaS
- 7. 7 Avoid Vendor Lock-in Open source software stack enables deployment on any public, private cloud or hybrid cloud
- 8. 8 Enable Unlimited Scalability Scales from several nodes on your laptop to tens of thousands of self-healing multi-tenant nodes
- 9. 9 Increase Agility and Maintainability By splitting applications into microservices with explicitly described dependencies
- 10. 10 Achieve Resiliency To failures of individual containers, machines, and even data centers and to varying levels of demand
- 11. 11 Improve Efficiency and Resource Utilization Via a central orchestrating process that dynamically manages and schedules microservices
- 12. Presentation received from Dan Played as it is âŠ.. for 20minutes.. CNCF Overview Dan Kohn, Executive Director, @dankohn1 This presentation is available at: https://github.com/cncf/presentations
- 15. 15 Quipu / Khipu a wearable data storage and communication system 500 to 5000 years ago cord color, manner of connection, relative placement, cord spacing, the types and placement of knots on individual cords https://en.wikipedia.org/wiki/Quipu Werenât we smartâŠ? We are becoming smarterâŠ.!
- 16. 16 We want to be smarter and smarterâŠ
- 17. 17
- 18. 18 Moving to Edge! Centralized to De-centralized to Distributed!
- 19. 19 Edge?
- 20. 20 Edge computing is a method of optimizing cloud computing systems by performing data processing at the edge of the network, near the source of the data.
- 21. 21 The possibility of intermittent cloud connectivity is a serious concern for mission-critical IoT applications such as a connected vehicle or other types of autonomous systems. âŠandâŠwhy Edge? Increasing costs of shipping the large volumes of data to the cloud for processing and storage. Data governance and security â many organizations have sensitive data that they donât want to leave their premises under any circumstances. Real-time decision making âthe latencies involved in shipping the data to the cloud for analytics are unacceptable. Reduce the Cost Trust & Security Real time, Ultra Low Latency Offline, Independent
- 22. 22 Roof Computing? Fog Computing? MEC ? Edge Computing?
- 23. 23 23 https://www.openfogconsortium.org IEEE 1931.1 Roof Computing WG
- 24. 24 The Markets and UsecasesâŠ
- 25. 25
- 26. 26 Some of the edge platforms⊠26 https://www.akraino.org/ Open Source Software Stack Supporting High-Availability Cloud Services Optimized For Edge The Open Platform for the IoT Edge StarlingX is a fully featured and high performance Edge Cloud software stack Specific Edge Platform https://wiki.opnfv.org/display/PROJ/Edge+cloud https://wiki.onap.org/display/DW/Edge+Automation+through+ONAP https://www.edgexfoundry.org/ https://wiki.openstack.org/wiki/StarlingX https://github.com/forestgiant An Open-Source Platform for Edge Computing
- 28. 28 Edge Node 1 Edge Node n D1 Dn D1 Dn Edge Node Edge Node 1 D1 Dn D1 Dn Central Cloud Edge Node n D1 Dn Sample Edge Deployment View âą Edge Clusters âą Scalable & Distributed âą North-South, East-West Communication âą Monitoring and Logging âą Security âą Different Resource Systems
- 29. 29 29
- 30. 30
- 31. 31
- 32. 32 One or more master nodes One or more worker nodes Distributed key-value store, like etcd https://kubernetes.io/docs/concepts/ Kubernetes (k8s) ArchitectureâŠ
- 33. 33 33 Edge Agent MS1 MS2 Edge Controller Edge Agent MS1 MS2 Edge Node Manager Edge Cloud Central Cloud Orchestration and Provisioning of Edge Nodes Orchestration and Provisioning of Edge Nodes Orchestration and Management of Microservices EN1 EN2 EN3 Each Microservice on container or pod Deployment Templates Deployment Templates
- 34. 34 34 Edge Agent MS1 MS2 EN2 MS2 Event Handl er Events pod CNI Edge Controller Edge Node Manager Edge Cloud Central Cloud EN1 Data Metrics and Monitoring Dashboard
- 35. 35 35 Edge Agent MS1 MS2 Edge Controller Edge Node Manager Edge Cloud Central Cloud EN1 EN2 EN3 Edge Bus Edge Bus Edge Agent MS1 MS2 Edge Bus Cloud Bus Edge Controller DNS Central DNS Edge DNS Edge DNS
- 36. 36 36 Edge Agent MS1 MS2 Edge Controller Edge Node Manager Edge Cloud Central Cloud EN1 EN2 EN3 Edge Bus Edge Bus Edge Agent MS1 MS2 Edge Bus Cloud Bus Workload Identity Management & Workload to workload security Spire Agent Spire Agent Spire Agent Infrastructure verification Server
- 37. 37 Cloud Cluster moving to Edge : Foster CNCF(1) Edge Agent MS1 MS2 Edge Controller Edge Node Manager Edge Cloud Central Cloud EN1 EN2 EN3 Edge Bus Edge Bus Edge Agent MS1 MS2 Edge Bus Cloud Bus Spire Agent Spire Agent Spire Agent Infrastructure verification Server Edge Controller DNS Central DNS Edge DNS Edge DNS
- 38. 38 Kubernetes : Orchestration and Management of Edge Nodes and or workloads Prometheus : Data Metrics and Monitoring Dashboard Helm : Deployment Charts Envoy: Proxy and Reverse Proxy for North-South and East-West Core DNS : DNS resolution and management (it helps for east west or north south DNS resolution) CNI : Workload networkings (container /pod) gRPC: Workload messaging NATS : Optimized Messaging service (esp for streaming data) for applications (pub/sub) Spiffe (spire) : workload to workload identity and authentication We can also haveâŠ.cloudevents for Serverless, rook for storage, opentracing for distributed tracing, opa for policy management.. Of course, rkt and conatinerd are part of our container runtime⊠âŠand more based on your deploymentâŠ. Cloud Cluster moving to Edge : Foster CNCF(2) Edge Cluster is a ~distributed Cloud ClusterâŠ
- 40. 40Thank You! Kumar Brothers! Krishna Kumar & Sanil Kumar Disclaimer: Images & many data are taken from Internet and only used for information sharing. We do not claim any other rights/correctness to it. Also no commercial usage of these slides allowed. âŠso as CNCF! Edge is CloudâŠ! Cloud be With you!