Git lab2019 jp-meetup
Download as PPTX, PDF0 likes53 views
The document discusses the "toolchain crisis" where different teams use unique sets of tools that cause integration complexity and slow down development. It notes how the shift to microservices has exacerbated this problem. It then summarizes how GitLab aims to be a single application that can handle the entire DevOps lifecycle, from planning to monitoring. It lists the many features GitLab has added over time to achieve this goal. The document concludes by stating GitLab's goal is to become the primary way to do DevOps on Kubernetes and provides a link to learn more about its roadmap.
Git lab2019 jp-meetup
- 2. 2 GitLab Product Vision&Direction in 2019
- 4. 4 The Toolchain Crisis Different teams are using unique set of tools and integrations
- 5. 5 Integration complexity of toolchains slows down teams https://about.gitlab.com/sdlc/#interfaces VerifyPlan CreateManage Package SecureMonitorConfigureRelease
- 6. 6 And it’s getting worse The shift to microservices creates an explosion of projects today’s toolchain is not built to handle.
- 7. 7 Microservices challenges monolith microservices ● Cloud computing and orchestration enables complexity ● This complexity manifests itself as: ○ Fragmentation ○ Data explosion ○ Siloed teams ○ Operational issues ○ Terrifying MTTR
- 8. 8 GitLab is the first single application for the entire DevOps lifecycle Manage Plan Create Verify Package Release Configure Monitor Secure Since 2016 GitLab added: Since 2011 GitLab added: Since 2011 GitLab added: Since 2012 GitLab added: Since 2016 GitLab added: Since 2016 GitLab added: Since 2018 GitLab added: Since 2016 GitLab added: Since 2017 GitLab added: Cycle Analytics DevOps Score Audit Management Authentication and Authorization Coming soon: Code Analytics Workflow Policies Kanban Boards Project Management Agile Portfolio Management Service Desk Coming soon: Value Stream Management Requirements Management Quality Management Source Code Management Code Review Wiki Snippets Web IDE Coming soon: Design Management Live Coding Continuous Integration (CI) Code Quality Performance Testing Coming soon: System Testing Usability Testing Accessibility Testing Compatibility Testing Container Registry Maven Repository Coming soon: NPM Registry Rubygem Registry Linux Package Registry Helm Chart Registry Dependency Proxy Continuous Delivery (CD) Release Orchestration Pages Review Apps Incremental Rollout Feature Flags Coming soon: Binary Authorization Auto DevOps Kubernetes Configuration ChatOps Coming soon: Serverless PaaS Chaos Engineering Runbook Configuration Cluster Cost Optimization Metrics Logging Cluster Monitoring Coming soon: Tracing Error Tracking Production Monitoring Incident Management Status Page SAST DAST Dependency Scanning Container Scanning License Management Coming soon: Web Application Firewall (WAF) RASP
- 9. 9 Software Composition Analysis Open source modules are checked to ensure proper licensing is in place. GitLab Recon Find new hosts to scan based on the top-level domain name. with Security Testing Automate security and compliancewith Portfolio Management Align strategy and execution Roadmaps & Epics Track progress and report on high-level status. Drill down into individual projects and issues to manage cross-project dependencies and distribute resources. GitLab Ultimate
- 10. 10 Become the way to do DevOps on K8s
Editor's Notes
- #5: The number of services and tools per services are both growing and each combination of these incurs a tax. The number of tools needed per project used to be just CI, Chef, Nagios for example - now, there are multiple categories. Time and cost to acquire all of these tools Time and cost to integrate all of these tools User context switching between all of these tools
- #6: Here you can see the core stages of the devops lifecycle, you have planning, code creation, all the way through to monitoring. And also shown, are are just some of the tools that can be potentially used at each stage. And this chaining together of tools, comes at the cost of integration complexity You need resources to design, build, and maintain the integrations, manage upgrades, and establish HA and DR for each one. They are brittle and can fail. Setting up a new project, or troubleshooting and optimizing an existing project, becomes difficult and is too slow for todays demands And the toolchain only serves to reinforce silios that you are trying to break down in the name of DevOps. Instead of teams being able to collaborate together, their different tools put them in indifferent workspaces.
- #7: As more and more development teams embrace Cloud Native (where applications are built using microservices running in containers), the number of projects proliferates which disparate tool sets were not built for. They were built during the monolithic application era, where releases encompassing a large number of changes were infrequent.