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

2015 State of DevOps Report

2015 State of DevOps Report

Nicole Forsgren (@nicolefv), Chef - For the fourth year running, the State of DevOps Report has looked into the trends and factors affecting DevOps transformations across industries and around the world. This year's report extends last year's work in investigating impacts on IT performance and organizational performance to look at things like:- burnout- lean management practices- continuous delivery- ... and more!

DevOpsDC Meetup

November 10, 2015
Tweet

More Decks by DevOpsDC Meetup

Other Decks in Technology

Transcript

  1. Some context… 2014 SoDO Report Scienced it up! Established a

    measure of IT Performance ◦ Throughput / agility: Deploy frequency, Lead Bme to deploy ◦ Stability / reliability: Mean Bme to restore. (Change fail rate is not included staBsBcally.) Discovered Three dis3nct groups of IT Performers: High, Medium, Low ◦ Analysis shows throughput and stability metrics move in tandem – effecBvely disproving ITIL claims that tradeoffs are necessary. Established a link between IT Performance and Org Performance ◦ Tying IT acBviBes and investments to profitability, market share, and producBvity is HUGE
  2. 2015 State of DevOps Report GOALS  Revalidate Core Findings from

    2014 SoDO Report ◦ IT Perf Measures sBll move in tandem ◦ Link between IT Performance and OrganizaBonal Performance  InvesBgate drivers of IT Performance – in a fully predic3ve model ◦ 2014 report largely reported correlaBons ◦ Expand to invesBgate impacts on how work *feels*  Look at demographics  InvesBgate a hunch in architecture…
  3. The Deets  4,976 respondents in 2015 ◦ 9,288 respondents in 2014

     Three groups of IT Performers: ◦ High ◦ Medium ◦ Low  Statistics BINGO: ◦ Correlation, CMV, PLS, SEM, EFA, varimax, Eigenvalue, factor, latent construct
  4. 2015: InvesNgate Drivers of IT Perf  DRIVERS means predic3on  One

    of three condiBons necessary for predicBve analysis to be appropriate ◦ Longitudinal analysis. NOPE ◦ Randomized study. NOPE ◦ Theory-based rela3onship. YES!  Based on expert opinion and a thorough review of the literature: ◦ ConBnuous delivery ◦ Lean Management
  5. One note  The rigor and transparency of the  2015 State

    of DevOps Report is impressive.  It was sBll a markeBng report, but it included numbers, details, citaBons (to work that was NOT from Puppet Labs) and even a brief methodology secBon.  Many thanks and congrats to Puppet Labs & IT RevoluBon Press
  6. Appendix: ApplicaNon Architecture Which of the following apply to the

    architecture of the primary applicaBon or service you are working on? (Select all that apply)
  7. Appendix: Services Architecture For services that your team has to

    interact with (new and exisBng), which of the following architecture abributes apply? (Select all that apply)