Dave: introduced the committee goal, scope, and deadline.
LF building a new CI as a service system as part of the CommunityBridge effort that will cut off support for our current Jenkins/Gerrit solution.
We should design a new system that not only gets all HL projects on to the same CI/CD platform but is also the same, or compatible with, the new LF system.
Ry: we need to get all of the CI/CD costs added up, including all of the money spent by outside companies.
HL should be paying for all of the CI/CD used by projects.
New system needs profiling capabilities.
Most important feature is that tests either succeed or fail and there is forensic capabilities in the CI/CD platform to quickly know if a failure is a legitimate failure in the software.
Shawn: must be driven by the PR process.
Both build passes as well as long-running tests.
Silas: plan to have public facing test nets for people to pound on and try to induce failures.
Action items
David Huseby move the initial CI/CD document from the TSC space to CI/CD space.
Shawn Amundson add details, including costs, of the Sawtooth and Grid CI/CD pipelines in the existing setups page.
Gregory Hill add details, including costs, of the Burrow CI/CD pipeline in the existing setups page.
David Huseby look into whether Kubernetes can utilize crowd-sourced computers to extend the cluster.
user-b2fb8 get the actual cost of the current Jenkins and Gerrit CI/CD setup used by Fabric.
David Huseby initial requirements gathering document set up.