Successfully reported this slideshow.
We use your LinkedIn profile and activity data to personalize ads and to show you more relevant ads. You can change your ad preferences anytime.

Relay: The Next Leg, Eric Sorenson, Puppet


Published on

Puppet Camp Australia

Published in: Technology
  • Be the first to comment

  • Be the first to like this

Relay: The Next Leg, Eric Sorenson, Puppet

  1. 1. Relay: The Next Leg Eric Sorenson // // @ahpook
  2. 2. /usr/bin/whoami sysadmin / sre / bad coder large scale config mgmt practitioner technical product manager guitarist and synthesist pinball game renter
  3. 3. 3
  4. 4. On-prem Cloud Cloud adoption has shifted the centre of gravity for automation • Event-driven orchestration • APIs, services, serverless, K8s, container platforms • Configuration management is a smaller part of the infra stack • Model/task-driven automation • OSs, hardware, apps, VMs, hypervisor • Configuration management is king Tomorrow Today
  5. 5. I need a complete solution to continuously deploy applications and infrastructure that provides: ● Simple deployment workflows built on using existing tools ● Support for multiple deployment targets ● Includes smart canaries, B/G deployments & automated rollback ● Lightweight governance and auditability ● Application deployment performance insights Nebula Project Nebula DEPLOY BUILD + TEST PROVISION ROLLBACK DEVOPS APP DEV GitOps CLI CI Existing CI tools Step Ecosystem
  6. 6. Cloud Native Resource Creation CN Resource Orchestration CN Workflow Automation Lyra endgame ● Combine forces with Nebula ● Contribute upstream to Tekton / CDF ● Integrate with other adjacent projects
  7. 7. Lesson 1: CD is Hard. Really Hard. source:
  8. 8. Lesson 2: Events Are Everywhere
  9. 9. Lesson 3: 'Ware The DevOps Dumping Ground source:
  10. 10. Relay is a system of intelligence that pulls together all the tools and technologies you need to effectively manage your environment. Secrets AI/ML/data Triggers Workflows K8s Serverless Cloud storage CI/CD Tools Config Mgmt Orchestrate actions across downstream tools and services Cloud events Listen to events from the services that you already use Incidents Tickets Git events Monitoring
  11. 11. Trigger steps are based on cloud events, git events, monitoring alerts, tickets, and incidents – like a Splunk alert, a Cloudwatch alert, a GitHub PR being merged, etc. Action steps are reusable, modular, and composable--things like destroying old EC2 instances, provisioning environments with Terraform, or clearing unattached EBS volumes. Query steps let you break out of a workflow to wait for an external process like manual approval or a callback from another system. Step Taxonomy
  12. 12. Workflows-as-Code YAML Parameters prompt the user for data input at runtime to customize the behaviour of the workflow. Values can be supplied interactively, via API, or on CLI. Secrets are stored securely server-side using Hashicorp Vault, referenced in a workflow using the "!Secret" tag. Output tags enable data from previous steps to be used by ones that come later.
  13. 13. Integration Ecosystem A library of useful integrations that cover the most common use cases and external services with sample workflows and well-documented steps that let you remix and combine however you want. Low-friction interoperability if you have existing containers that you want to use in Relay workflows, they can drop right in alongside purpose-built integrations. Easy step authoring if you've built something that other people might find useful or want to contribute to existing steps
  14. 14. 14 Coming Real Soon Now™ (early Northern Hemisphere summer) Sign up for the notification list! Join us in Slack for Q&A and early access: #relay Thank you! Eric Sorenson // // @ahpook