#DataDrivenDevOps

A presentation at Cincy Deliver 2019 in July 2019 in Cincinnati, OH, USA by Baruch Sadogursky

Slide 1

Slide 1

#DataDrivenDevOps ”Without data you are just another person with an opinion” W. Edwards Deming @jbaruch #CincyDeliver #DataDrivenDevOps http://jfrog.com/shownotes

Slide 2

Slide 2

shownotes http://jfrog.com/shownotes Slides Video Links Comments, Ratings Raffle @jbaruch #CincyDeliver #DataDrivenDevOps http://jfrog.com/shownotes

Slide 3

Slide 3

Cobbler’s children have no shoes

Slide 4

Slide 4

Poll time! How many of you are software engineers? How many of you are optimists? How many of you are self-confident in their work? @jbaruch #CincyDeliver #DataDrivenDevOps http://jfrog.com/shownotes

Slide 5

Slide 5

Dunning-Kruger Effect a.k.a. ”optimism” “ People suffer from illusory superiority, mistakenly assessing their cognitive ability as greater than it is. Wikipedia @jbaruch #CincyDeliver #DataDrivenDevOps http://jfrog.com/shownotes

Slide 6

Slide 6

Second-system effect a.k.a. “self-confidence” “ The tendency of small, elegant, and successful systems, to be succeeded by overengineered, bloated systems, due to inflated expectations and overconfidence. The Mythical Man-Month @jbaruch #CincyDeliver #DataDrivenDevOps http://jfrog.com/shownotes

Slide 7

Slide 7

The consequences @jbaruch #CincyDeliver #DataDrivenDevOps http://jfrog.com/shownotes

Slide 8

Slide 8

disillusion Software I like @jbaruch #CincyDeliver #DataDrivenDevOps Software I know really well http://jfrog.com/shownotes

Slide 9

Slide 9

overconfidence “ Every software is 80% ready 80% of the time Every developer, ever @jbaruch #CincyDeliver #DataDrivenDevOps http://jfrog.com/shownotes

Slide 10

Slide 10

Hard to give up @jbaruch #CincyDeliver #DataDrivenDevOps http://jfrog.com/shownotes

Slide 11

Slide 11

@jbaruch #CincyDeliver #DataDrivenDevOps http://jfrog.com/shownotes

Slide 12

Slide 12

How departments measure Support Marketing Sales HR Finance @jbaruch #CincyDeliver Engineering Anecdotes from daily standup #DataDrivenDevOps http://jfrog.com/shownotes

Slide 13

Slide 13

@jbaruch #CincyDeliver #DataDrivenDevOps http://jfrog.com/shownotes

Slide 14

Slide 14

Slide 15

Slide 15

🎩 @jbaruch #dockercon jfrog.com/shownotes @ErinMeyerINSEAD’s “Culture Map”

Slide 16

Slide 16

Engineering metrics in pre-devops era Ops Dev • Velocity • Architecture • Compliance @jbaruch #CincyDeliver • SLA • Cost env setup • Average customer cost • Tools stability #DataDrivenDevOps • • • • • QA Incidents Defects MTTR Code coverage Test suite stability http://jfrog.com/shownotes

Slide 17

Slide 17

DEVOPS! ? @jbaruch #CincyDeliver #DataDrivenDevOps http://jfrog.com/shownotes

Slide 18

Slide 18

Velocity to agile is like ? To devops… ? @jbaruch #CincyDeliver #DataDrivenDevOps http://jfrog.com/shownotes

Slide 19

Slide 19

Velocity to agile is like Flow Efficiency To devops… Flow Efficiency @jbaruch #CincyDeliver #DataDrivenDevOps http://jfrog.com/shownotes

Slide 20

Slide 20

Very much alike! Metric Flow Efficiency Velocity @jbaruch Easily understandable Unity actionable ✓ ✓ ✗ ✓ ✓ ✗ #CincyDeliver #DataDrivenDevOps http://jfrog.com/shownotes

Slide 21

Slide 21

Poll time! Who knows what velocity is? Who knows what burndown chart is? Who has a burndown chart? Who looks at the burndown chart? Who trusts the burndown chart? Who knows what to do if it doesn’t look right?

Slide 22

Slide 22

What’s the problem? Velocity @jbaruch #CincyDeliver #DataDrivenDevOps http://jfrog.com/shownotes

Slide 23

Slide 23

True for both! Metric Flow Efficiency Velocity @jbaruch Easily understandable Unity actionable ✓ ✓ ✗ ✓ ✓ ✗ #CincyDeliver #DataDrivenDevOps http://jfrog.com/shownotes

Slide 24

Slide 24

Too coarse! Flow Efficiency @jbaruch #CincyDeliver #DataDrivenDevOps http://jfrog.com/shownotes

Slide 25

Slide 25

Too coarse! DevOps @jbaruch #CincyDeliver #DataDrivenDevOps http://jfrog.com/shownotes

Slide 26

Slide 26

This is devops* Ops Deep specialization Dev *Unless you’re netfliX **You’re not @jbaruch #CincyDeliver Common goals, tools, culture QA #DataDrivenDevOps http://jfrog.com/shownotes

Slide 27

Slide 27

This is devops Ops Dev @jbaruch #CincyDeliver QA #DataDrivenDevOps http://jfrog.com/shownotes

Slide 28

Slide 28

Same for metrics! Ops Specialized metrics Dev @jbaruch #CincyDeliver Flow Efficiency QA #DataDrivenDevOps http://jfrog.com/shownotes

Slide 29

Slide 29

Same for metrics! Ops Dev @jbaruch #CincyDeliver QA #DataDrivenDevOps http://jfrog.com/shownotes

Slide 30

Slide 30

How do metrics collaborate? influenced affected dev dev qa • Time to build • False reopen count • Release pipelines • QA coverage stability • Artifact replication topology ops • qa • • @jbaruch ops Avg cost of customer • Smoke tests quality Test suite stability • Time to create new Incoming defect QA lab velocity http://jfrog.com/shownotes • #CincyDeliver Test execution#DataDrivenDevOps time

Slide 31

Slide 31

Cooking with metrics @jbaruch #CincyDeliver #DataDrivenDevOps http://jfrog.com/shownotes

Slide 32

Slide 32

Metrics categorization @jbaruch #CincyDeliver #DataDrivenDevOps http://jfrog.com/shownotes

Slide 33

Slide 33

Why? @jbaruch #CincyDeliver #DataDrivenDevOps http://jfrog.com/shownotes

Slide 34

Slide 34

ROI – is this activity is the right thing to do?

Slide 35

Slide 35

Continuous improvement

Slide 36

Slide 36

trust

Slide 37

Slide 37

sampling @jbaruch #CincyDeliver #DataDrivenDevOps http://jfrog.com/shownotes

Slide 38

Slide 38

Complete sample

Slide 39

Slide 39

@jbaruch #CincyDeliver #DataDrivenDevOps http://jfrog.com/shownotes

Slide 40

Slide 40

Representative sample

Slide 41

Slide 41

How effective are we? Samples are good enough for that As long as they are representative Need to be collected over time @jbaruch #CincyDeliver #DataDrivenDevOps http://jfrog.com/shownotes

Slide 42

Slide 42

Predicting the future

Slide 43

Slide 43

Meet scrum team “avengers”

Slide 44

Slide 44

anecdotes from the daily scrum stand-ups “ We never get enough testing environments from Ops, QA is suffering. Scrum team “Avengers” @jbaruch #CincyDeliver #DataDrivenDevOps http://jfrog.com/shownotes

Slide 45

Slide 45

Let’s gather some data! Environment Utilization 10 8 6 4 2 0 week 1 week 2 week 3 week 4 week 5 week 6 week 7 Utilized @jbaruch #CincyDeliver week 8 week 9 week 10 week 11 week 12 week 13 week 14 Available #DataDrivenDevOps http://jfrog.com/shownotes

Slide 46

Slide 46

That doesn’t look good! But why? Days the environment stays checked out 25 25 20 18 15 13 13 John Scott 14 15 16 10 10 7 6 5 1 2 3 0 Raj Jane Christina Helen Lee Baruch Nataly Ruppert Josh Patrick Michael Days checked out @jbaruch #CincyDeliver #DataDrivenDevOps http://jfrog.com/shownotes

Slide 47

Slide 47

Environment Utilization 16 14 12 Continuous improvement 10 8 6 4 2 0 week 1 week 3 week 5 week 7 week 9 week 11 week 13 week 15 week 17 week 19 week 21 week 23 Utilized @jbaruch #CincyDeliver #DataDrivenDevOps Available http://jfrog.com/shownotes

Slide 48

Slide 48

anecdotes from the daily scrum stand-ups “ Tests are failing, because the test suite is not stable, QA makes our builds fail without reason. Scrum team “Avengers” @jbaruch #CincyDeliver #DataDrivenDevOps http://jfrog.com/shownotes

Slide 49

Slide 49

Let’s gather some data!

Slide 50

Slide 50

Return on investment @jbaruch #CincyDeliver #DataDrivenDevOps http://jfrog.com/shownotes

Slide 51

Slide 51

anecdotes from the daily scrum stand-ups “ Those developers want to deploy new GC, its tons of work, without any profit. Scrum team “Avengers” @jbaruch #CincyDeliver #DataDrivenDevOps http://jfrog.com/shownotes

Slide 52

Slide 52

Let’s gather some data! @jbaruch #CincyDeliver #DataDrivenDevOps http://jfrog.com/shownotes

Slide 53

Slide 53

0:00 0:40 1:20 2:00 2:40 3:20 4:00 4:40 5:20 6:00 6:40 7:20 8:00 8:40 9:20 10:00 10:40 11:20 12:00 12:40 13:20 14:00 14:40 15:20 16:00 16:40 17:20 18:00 18:40 19:20 20:00 20:40 21:20 22:00 22:40 23:20 GC time Trust and predictability GC time @jbaruch #CincyDeliver #DataDrivenDevOps http://jfrog.com/shownotes

Slide 54

Slide 54

Dos and Don’ts

Slide 55

Slide 55

Measuring the right thing

Slide 56

Slide 56

LoC LOL, anyone? @jbaruch #CincyDeliver #DataDrivenDevOps http://jfrog.com/shownotes

Slide 57

Slide 57

Avoid data overload

Slide 58

Slide 58

Use reliable data

Slide 59

Slide 59

”medium” In brazil ”medium” In russia Common vocabulary is important

Slide 60

Slide 60

Metrics generate incentives, be aware! @jbaruch #CincyDeliver #DataDrivenDevOps http://jfrog.com/shownotes

Slide 61

Slide 61

Tweak as you go

Slide 62

Slide 62

Elevate! Integrated Dashboards Product Specific Dashboards Raw Data @jbaruch #CincyDeliver #DataDrivenDevOps http://jfrog.com/shownotes

Slide 63

Slide 63

@jbaruch #CincyDeliver #DataDrivenDevOps http://jfrog.com/shownotes

Slide 64

Slide 64

How departments measure Support Marketing Sales HR Finance @jbaruch #CincyDeliver Engineering Anecdotes from daily standup #DataDrivenDevOps http://jfrog.com/shownotes

Slide 65

Slide 65

Step into data-driven life Removes blame game Builds accountability and trust Creates common base for discussion Enables DevOps @jbaruch #CincyDeliver #DataDrivenDevOps http://jfrog.com/shownotes

Slide 66

Slide 66

Q&A and twitter ads @jbaruch #CincyDeliver #DataDrivenDevOps http://jfrog.com/shownotes