Attendees:
Topics:
- Gap Analysis: focus on Releng Support for OPNFV
- Overview by Trevor Bramwell
Notes:
- Overview Releng
- Group within LF & LFN
- Release Engineering support LF and LFN
- See Trevor’s link
- Environments provided to customers
- Best practices
- Training
- Consulting
- Helping project maintain jobs: e.g. Jenkins and other CI systems
- Support queue that provide ticket and support for building and managing CI
- Don’t do user support (i.e. users of project softwares)
- End-User Advisory Groups are used for operators and users of software. ONAP has one and ONAP has one, which
- Exists to get feedback from end user stakeholders.
- Typical tickets: job failed, external system, caching, new project onboarding, etc. standing up project infrastructure solutions
- Most of work has been: “hey can you please release X artifact”, runs process to push out (ONAP has automated a lot of this)
- New work has been mostly supporting infrastructure automation systems.
- Answer community questions: reqs for systems, options available (e.g. security scanning), helps community,
- More than releng Trevors group is also greater technical liason for LF & LFN. When LFN umbrella happened. More streamlining was hoped for but that doesn’t scale so they work as a ‘group’ / pool.
- Modernization WIP to streamline/collaborate/etc. – has looked at reqs, tools, still trying to figure out CI. Also trying to continue moving
- Towards self-service, automate more and more of what being done.
- **Having pipelines in same repos helps this** but right now each project has it’s own repo. Makes for a handoff to leverage CI.
- In an LF project, have transitioned from Jenkins to ‘Azure pipeline. Project is given tools and access to make changes themselves. Trevor’s team th
- Then work as advisory role. Always looking for leverage and scale, e.g. Global JJB work new Java project can get up and running fast with templates
- Predefined and ready to go.
- Tools
- Continuous Integration Circle CI, Jenkins
- Continuous Delivery (of artifacts): Nexus, S3, Artifactory Cloud, Jfrog bintry, DockerHub, Maven, PyPI, etc. – different per project
- Continuous Deployment: generally do not do this part
- Currently OPNFV using Jenkins and GoogleCompute and DockerHub. On Gerritt for code. OPNFV has a lot of hardware, but need to more of cloud access.
- Session in Prague. What is needed for CNTT CI systems?
- Orange folks at ONAP are evaluating GitLab CI, and evaluating K8s as a service. They are currently working with Azure. Cedric is close affiliated with the ONAP work
- So can/may trickle down to ONPFNV.
- Trevor’s group has 8 folks right now, with a few contractors. For all intents and purposes Trevor and Aric are the SPOCs for OPNFV, and both will be in Prague.
- Jessica Wagantahl. Bengt Thuree. Will also be in Prague.