Upgrade to Pro — share decks privately, control downloads, hide ads and more …

Lessons learned from 7 years of running develop...

Coté
October 02, 2023

Lessons learned from 7 years of running developer platforms

This talk covers best and worst practices for Platform engineering. The trick is, we didn’t always use this phrase, so we actually have many years of experience to learn from. Who are you? You’re in a DevOps, wait, I mean SRE…nope, scatch that…“platform engineering” team. People are coming at you to get kubernetes up and running and then build some kind of platform on-top of kubernetes. But you just got a build pipeline in place! Getting kubernetes ready for developers may be a new problem, but building and running developer platforms has been going on for at least ten years. This talk will cover the lessons those organizations have learned such as: product managing the platform, attracting and retaining developers, seeding trust and skills, re-skilling existing ops staff, and more. Examples are drawn from organizations like Mercedes-Benz, the US Airforce, large insurance companies and banks, and more.

See my platform page for more platform related stuff.

Coté

October 02, 2023
Tweet

More Decks by Coté

Other Decks in Technology

Transcript

  1. © VMware, Inc. 7 Lessons from 7 Years of Running

    Platforms Or, “what is platform engineering.” Coté October, 2023
  2. © VMware, Inc. 2 Sources: see “How's DevOps been going?,”

    Coté, June 2023 for citations and links to sources. Accounts of deployment rates vary wildly 81% 65% 42% 26% DORA (2022) CD Foundation (2023) Forrester (2021) Forrester (2022) Deploy Monthly or Less
  3. © VMware, Inc. 5 Source: "Improving JPMorgan Chase's Developer Experience

    on the Cloud," Nadi Awad, July, 2022. Lessons learned from supporting 1,500+ application at JP Morgan Chase
  4. © VMware, Inc. 6 Sources: CNCF Platforms White Paper and

    Oct 2022 diagram draft; VMware Tanzu Application Platform. 1
  5. © VMware, Inc. 7 Platforms == Internal Developer Portals +

    CaaS …more or less… Gartner’s IDP model VMware Tanzu Platform Model Source: Gartner, Innovation Insight for Internal Developer Portals, By Manjunath Bhat, Mark O'Neill, Oleksandr Matvitskyy 1 February 2022. 3
  6. 8 © VMware, Inc. “We are building this platform not

    for us, we are building it for Mercedes-Benz developers.” Thomas Müller, Mercedes-Benz 2
  7. © VMware, Inc. 10 Reactive, be-spoke, repetitive Manual ticket driven,

    built based on customer demand Automated, self-service platforms with well defined products Increasing FTE Creating champions within own customer base Service Delivery Platform Approach Considered as Cost and Bottle Neck Considered as Experts that add Value KPIs / OKRs based on uptime / capacity KPIs / OKRs based on customer engagement and value Success Measurement Customer Interactions Demand Management Reputation How to Scale Reputation Service reviews, roadmaps, feedback loops, documentation Transform the ops culture Source: Darran Rice (formally at UBS) in “Improve Developer Productivity with Platform as a Product,” VMware Explore, Nov. 2022.
  8. © VMware, Inc. 13 Find the Developer Toil, Confusion, Blockers

    • What are we making? • We have a strong vision for our product, and we're doing important work together every day to fulfill that vision. • I have the context I need to confidently make changes while I'm working. • I am proud of the work I have delivered so far for our product. • I am learning things that I look forward to applying to future products. • My workstation seems to disappear out from under me while I'm working. • It's easy to get my workstation into the state I need to develop our product. • What aspect of our workstation setup is painful? • It's easy to run our software on my workstation while I’m developing it. • I can boot our software up into the state I need with minimal effort. • What aspect of running our software locally is painful? What could we do to make it less painful? • It's easy to run our test suites and to author new ones. • Tests are a stable, reliable, seamless part of my workflow. • Test failures give me the feedback I need on the code I am writing. • What aspect of production support is painful? • We collaborate well with the teams whose software we integrate with. • When necessary, it is within my power to request timely changes from other teams. • I have the resources I need to test and code confidently against other teams' integration points. • What aspect of integrating with other teams is painful? • I'm rarely impacted by breaking changes from other tracks of work. • We almost always catch broken tests and code before they're merged in. • What aspect of committing changes is painful? • Our release process (CI/CD) from source control to our story acceptance environment is fully automated. • If the release process (CI/CD) fails, I'm confident something is truly wrong, and I know I'll be able to track down the problem. • Our team releases new versions of our software as often as the business needs us to. • We are meeting our service-level agreements with a minimum of unplanned work. • When something is wrong in production, we reproduce and solve the problem in a lower environment. Sources: "Developer Toil: The Hidden Tech Debt," Susie Forbath, Tyson McNulty, and Coté, August, 2022. See also Michael Galloway’s interview questions for platform product managers. 4
  9. © VMware, Inc. 14 Sources: BT Canvas team; MB.io; Duke

    Energy; Allstate; "Take DevOps to 11 and Sprinkle Cloud on it with Rainbows and Unicorns," Matt Curry, s1p 2017. “Improve Developer Productivity with Platform as a Product,” VMware Explore, Nov. 2022. Platform marketing, advocacy, consulting 5
  10. © VMware, Inc. 15 Initial Phase – Create the Platform

    with Developers 1. Platform team: owner, operators, engineers, advocate 2. Pick one app for business & technical feasibility. 3. Developer toil audit. 4. Find path to production with end-to-end value stream analysis 5. Start with pre-integrated platform, customize as you... 6. Build a golden path with the developer team. 7. Optionally, build an IDP as part of the platform. 8. Do this for 3 months. 6
  11. © VMware, Inc. 16 Scaling Phase: Pairing & Seeding to

    build trust & training 1. Create platform marketing program. 2. Find two to five more apps. 3. Pair & seed from first dev & platform team to new teams. 4. "Shift Left" - build golden paths for governance, security, etc. 5. Add more infrastructure staff with pairing & seeding. 6. Do this for three months. 7. Repeat, growing number of apps as pairing & seeding allows. 7 Sources: “From 0 to 1000 Apps: The First Year of Cloud Foundry at The Home Depot,” Anthony McCulley, The Home Depot, Aug 2016; “Cloud Native at The Home Depot, with Tony McCulley,” Pivotal Conversations #45; USAF presentations and write-ups; "Driving Business Agility Without Large-Scale Transformation Programs," Venkatesh Arunachalam, Sep 2021; The Home Depot 2022[?]Q4 earnings call; The Business Bottleneck, Coté.
  12. © VMware, Inc. 17 All too common challenges & blockers

    People & Culture • ITSM instead of Platforms • Overwhelming legacy portfolio • Dependencies between teams • Local optimization, no CI/CD • Budgeting • Misaligned executives • IT is still in the basement • Compliance • Weak connection to business value Technical Execution • Skills, hiring • Reluctance to change • Scaling new roles • Org. structure • “We already do agile.” • Sustaining change Plans & Alignment