Evolving Team Structure in DevOps

Evolving Team Structure in DevOpsSherry Chang

Intel IT Chief Architect, DevOps & Continuous Delivery

https://www.linkedin.com/in/sherryschang

Legal NoticesThis presentation is for informational purposes only. INTEL MAKES NO WARRANTIES, EXPRESS OR IMPLIED, IN THIS SUMMARY.

[include if relevant to your presentation] Software and workloads used in performance tests may have been optimized for performance only on Intel microprocessors. Performance tests, such as SYSmark and MobileMark, are measured using specific computer systems, components, software, operations and functions. Any change to any of those factors may cause the results to vary. You should consult other information and performance tests to assist you in fully evaluating your contemplated purchases, including the performance of that product when combined with other products.

For more complete information about performance and benchmark results, visit www.intel.com/benchmarks

Intel and the Intel logo are trademarks of Intel Corporation in the U.S. and/or other countries. [insert other relevant trademarks used in this presentation. For a list of Intel trademarks, go to http://legal.intel.com/Trademarks/NamesDb.htm]

* Other names and brands may be claimed as the property of others.

Copyright © 2015, Intel Corporation. All rights reserved.

2

http://www.intel.com/benchmarks

http://legal.intel.com/Trademarks/NamesDb.htm

3

About Me

• Chief Architect, DevOps & Continuous Delivery– Working at Intel IT

• 20+ Years in Software Development• ITIL V2 Certified• Certified Scrum Master• BBA, Management Information

Systems– University of Texas at Austin

BiographySherry Chang

2015 Intel IT Vital Statistics>6,065 IT employees

50 global IT sites

>106,000 Intel employees1

170 Intel sites in 66 Countries

61 Data Centers(91 Data Centers in 2010)80% of servers virtualized(42% virtualized in 2010)

>119,000+ Devices100% of laptops encrypted100% of laptops with SSDs>53,700 handheld devices

164 mobile applications developed

4Copyright © 2015, Intel Corporation. All rights reserved.

5

Questions from Leaders– Do we need a re-organization to

better align?– What is the “perfect” DevOps

team?

6

TEAM STRUCTURE EVOLUTION“Organizations which design systems… are constrained to produce designs which are copies of the communication structures of these organizations”

-- M. Conway

7

Waterfall

BusinessAnalyst

Tech Analyst

User Experience

Dev QA/ Test

OpsStakeholder

Project

Value StreamInception Delivery

Analysis Design Coding Testing Release & Operate

Silo of Functional Specialists

8

Agile Transformation

Stakeholder

Developer

QA/Test

Developer Developer

TeamLead

Cross Functional Team Delivering Value Sooner

9

Agile: What Actually Happens

Stakeholder

Developer

QA/Test

Developer Developer

TeamLead

Dev: Agile Team for Projects

Ops: ITSM Team for KTBR

Customers

Problem Manager

Incident Manager

Capacity Manager

Release Manager

Service Manager

Dev & Ops Silo


10

Ideal DevOps Team Structure

Stakeholder

Developer

QA/Test

Developer SysAdmin

TeamLead

Services

Stakeholder

Developer

QA/Test

Developer SysAdmin

TeamLead

Services

API API

Self Service

Team 1 Team 2

•Small (5-9 people)•Persistent•Self-Organizing•Cross-Functional,

•Loosely Coupled •Owns 1 or more services (or micro-services)•All provided services are Self-Service•All provided services have API

11

However..

Stakeholder

Developer

QA/Test

Developer SysAdmin

TeamLead

Service

What if Service Does Not Fit 2 Pizza Team?


12

No one size fits all.. But guidance available for designing the right structure for you.

13

DESIGNING THE RIGHT STRUCTURE

14

Guidance #1 – Optimize Handoff

BusinessAnalyst

Tech Analyst

User Experience

Dev QA/ Test

OpsStakeholder

Project

XX X X X

Business Analyst

API

Self Service A/B Testing

TestEngineers

API

Self Service Test Automation Framework

API

Feature Developers

API

Self Service Features

API Operation

API

Self Service Platform

API

Provide Function as Self Service

Eliminate Tickets Eliminate Wait on

Resource Availability

Provide API to Enable Automation

15

Guidance #2

First understand your value stream..


WorkWait

Request

New Feature Requests

Approve & Prioritize

Size & Prioritize Code Test Deploy

Perform By:Customers

Perform By:Portfolio Manager

Perform By:Dev Team

Perform By:Dev Team

Perform By:Test Team

Perform By:OpsTeam

Submit Ticket

Assign Team Schedule

Assign Tester

Wait for Ops

30 min3 Days

30 min15 Days

15 min15 Days

2 Hours2 Weeks

2 Hours2 Weeks

8 Hours

15 min2 Days

2 Hours2 Days

4 Hours

Total19.5 Hours9 Weeks

ManualManual

16

Guidance #2Then Optimize It!


WorkWait

New Feature Requests

Story to Backlog

Size & Prioritize Code Test Deploy

Perform By:PO

Perform By:Agile Team



Perform By:OpsTeam

Assign Team Schedule

Assign Tester

Click Button

30 min15 Days

2 Hours2 Weeks

2 Hours

8 Hours

20 min

20 min

Total13 Hours4 Weeks

Automate

Automate

XX

Improvement Opportunities• Eliminate Handoff• Eliminate Work• Automate Work• Combine Multiple Streams of Work

17

Emergency Changes

Standard Changes

Normal Changes

Major Changes

Continuous Delivery Platform

Consolidating Multiple Streams

Release

18

Guidance #3

Amplify Feedback Loop Thru• Shared Tools• Shared Pipeline• Shared Learning

Feedback Loop

Unit Test CI

BuildSyste

m

Regression

Production

Monitoring

Customer

Feedback

DevQAOps

Dev Ownership

QA Ownership

Ops Ownership

19

Ingredient

Integration

Ingredient

Regression

System Integration

System

Regression

Release Validation

Ingredient(SW/FW)

Ingredient(SW/FW)

Ingredient (OS)

Ingredient(TPV)

Regression Test

RegressionTest

Continuous System

Integration

IngredientValidation

Ingredient Validation

System Validation

System Regression

Example Structure: Feedback Loop in Action

1 Hour or Less

4-6 Hours

1 Hour or Less

4-6 Hours 1

WeekFeedback Loop

20

Guidance #4 – Codify Change Control Board (CCB)

Stakeholder

Developer

QA/Test

Developer Developer

TeamLead

Dev Ops

Customers

Problem Manager

Incident Manage

r

Capacity Manage

r

Release Manage

r

Service Manage

r


XCCB


Code Pre-CommitValidation

Build & Integrate Quality Gates Deployment

Automated Automated Automated Manual Automated

CCB

Codified Policy

21

Expand Codify Policy


CCB


Code Pre-CommitValidation

Build & Integrate Quality Gates Deployment

Automated Automated Automated Manual Automated

CCB

Codified Policy

Legal

Intellectual Property Scan

Governance

BlacklistTechnologyScan

Security

VulnerabilityScan

Advantages Speed Consistent Application Audit Trail

22

TRANSFORMING THE ORGANIZATION

• Manual• Reliance on documented steps• Tickets for service requests

• Manual testing after development• Little or no automation• Lack quality measures

• Primarily human gates• Policy captured in documents• Process designed for infrequent

releases

Current

• Developer skills to codify infrastructure Developer skills to automate deployment• Develop self service infrastructure

• Test engineer skills to automate tests• Develop quality metrics measure and

tracking• Behaviorl Driven Development

• Develop self service automated quality gate• Codify policy• Develop self service metrics

To Be

• Manual tests after development• Big bang integration • Manual build and deployment

• Test Driven Development• Continuous Delivery• System Administration skills to consume

infrastructure service

Development

Operation

Testing

Governance

Skill Transformation

24

DevOps Culture

• Open Source (Internal or External)• Social Collaboration • Experiment, A/B Testing• Metric Driven Decision (Minimize PPT, Plan, etc.)

25

Conclusion• No “Perfect” DevOps Teams• Structure Should not be the First Concern• Do Not Need a Re-Org• Focus on Eliminating Friction and Wait Between Teams

26

Resources I Found ValuableFor Value Stream

AnalysisFor Amplified Feedback

Loop

For Ideal Team Structure

http://www.slideshare.net/ertansonat/spotify-engineering-culture-52284519

27

QUESTIONS?

Thank you for attending this session.

Please don’t forget to complete a session evaluation on the conference app.

Evolving Team Structure in DevOps

Technology

Transcript of Evolving Team Structure in DevOps