Google Summer of Code and Layer5 Community

      What is Layer5 Community?

      The Layer5 community represents the largest collection of cloud native management integrations. We build projects to provide learning environments, deployment and operational best practices, performance benchmarks, create documentation, share networking opportunities, and more. Our shared commitment to the open source spirit pushes Layer5 projects forward. New members are always welcome.

      Is it Open Source?

      Layer5 projects are open source software. Anyone can download, use, work on, and share it with others. It's built on principles like collaboration, globalism, and innovation. Layer5 projects are distributed under the terms of Apache v2.

      Google Summer of Code Participation?

      The key component of these projects is our Community. This community, which you will join as an participant in Google Summer of Code, is improving the world of diverse cloud native systems. Your contributions will affect people you've never met. The Layer5 community includes software engineers, researchers, students, artists, system administrators, operators and web designers -- all of whom will be happy to help you get started.
      We believe that all contributors should expect and be part of a safe and friendly environment for constructive contribution. We can more effectively and successfully compare and challenge different ideas to find the best solutions for advancement, while building the size, diversity, and strength of our community.

2020 Program Timeline

  • - January 14 - Organization applications open
  • - February 20 - Accepted GSoC Organizations announced
  • - March 16-March 31 - Students submit their proposals
  • - April 27 - Accepted students are announced
  • - April 27-May 18 - Community bonding period with orgs
  • - May 18-Aug 17 - Students code the summer away
  • - August 25 - Successful student projects are announced

Statistics

  • - In 15 years 15,926 students from 109 countries have been accepted into GSoC
  • - Countries with the most students: India (4,065), United States (2,552), and Germany (870 )
  • - Approximately 36+ million lines of code have been produced
GSOC 2020 Project Ideas

      SMI Conformance Testing

      Description

      : Ensure that a service mesh is properly configured and that its behavior conforms to official SMI specifications. Conformance consists of both capabilities and compliance status as outlined in the design specification.

      Distributed Load Testing of Service Meshes

      Description

      : Many performance benchmarks are limited to single instance load generation (single pod load generator). This limits the amount of traffic that can be generated to output of the single machine that the benchmark tool runs on in or out of a cluster. Overcoming this limitation would allow for more flexible and robust testing. Distributed load testing in parallel poses a challenge when merging results without losing the precision we need to gain insight into the high tail percentiles. Use Meshery’s interface for supporting different load generators (a point of extensibility) to place support for Vegeta. Workloads running on service meshes are hit in various ways and generally not from the same source. Use of a distributed load generator will aid in the testing of service mesh behavior.

Recommended Skills:

Golang, Kubernetes

Mentors:

Issue:

https://github.com/servicemeshinterface/smi-spec/issues/70
https://github.com/envoyproxy/envoy-perf/issues/72
GSOC 2019 Project Ideas

      Meshery

      Meshery is the open source, cloud native manager that enables the adoption, operation, and management of Kubernetes, any service mesh, and their workloads. See Meshery for more information.

      Linkerd and Envoy

      Linkerd is an ultralight service mesh for Kubernetes and beyond: https://linkerd.io. Envoy is an open source edge and service proxy, designed for cloud-native applications: https://www.envoyproxy.io.

      Benchmarks for Linkerd and Envoy

      Description: Linkerd, like other service meshes are plagued by the question of adopters asking the question: "what's the performance overhead of the service mesh?". Envoy does not publish performance test results (see How fast is Envoy). Linkerd, Istio, Envoy and the list of other service meshes don't have a consistent set of performance benchmarks between them. So, even if Envoy were to publish performance results, users still wouldn't be able to compare overhead between Linkerd and Envoy. The project idea here is to build a multi-mesh performance benchmark tool.

Recommended Skills:

Golang, JavaScript, Kubernetes

Mentors:

Issue:

https://github.com/envoyproxy/envoy/issues/5536
https://discourse.linkerd.io/t/linkerd-performance/146