the data on devops: making the case for awesome

45
Nicole Forsgren, PhD Director of Organizational Performance & Analytics, Chef Research Affiliate, Clemson University The Data on DevOps: Making the Case for Awesome

Upload: nicole-forsgren

Post on 14-Apr-2017

1.722 views

Category:

Technology


1 download

TRANSCRIPT

Page 1: The Data on DevOps: Making the Case for Awesome

Nicole Forsgren, PhD

Director of Organizational Performance & Analytics, ChefResearch Affiliate, Clemson University

The Data on DevOps:Making the Case for Awesome

Page 2: The Data on DevOps: Making the Case for Awesome

“IT doesn’t matter.”-- Nicholas Carr, 2003

@nicolefv

Page 3: The Data on DevOps: Making the Case for Awesome

IT Does Matter

• Times – and IT – have changed• DevOps is good for Organizations• DevOps is good for IT• And then some detail: What drives this change?

• Technical practices (hint: Continuous Delivery)• Management practices (hint: from Lean Management)• Culture and identity

@nicolefv

Page 4: The Data on DevOps: Making the Case for Awesome

@nicolefv

Page 5: The Data on DevOps: Making the Case for Awesome

@nicolefv

Page 6: The Data on DevOps: Making the Case for Awesome

DevOps isTechnical practices

seen in Continuous Delivery,Management practices

seen in Lean Management principles, and Organizational Culture and Identity

@nicolefv

Research shows that these drive IT Performanceand Organizational Performance

Page 7: The Data on DevOps: Making the Case for Awesome

DevOps is good for organizations

@nicolefv

Page 8: The Data on DevOps: Making the Case for Awesome

High Performing IT organizations

2xMore likely to exceed Profitability, Market share, andProductivity goals

50%Higher market capgrowth over 3 years*

The 2014 and 2015 State of DevOps Report

@nicolefv

Page 9: The Data on DevOps: Making the Case for Awesome

Devops is good for IT

Measuring DevOps and IT Performance- Deploy frequency (Note: NOT delivery)- Lead Time for Changes- Mean Time to Recover (MTTR)- Change Fail Rate

@nicolefv

Page 10: The Data on DevOps: Making the Case for Awesome

High Performing DevOps teams

More agile

The 2016 State of DevOps Report

@nicolefv

Page 11: The Data on DevOps: Making the Case for Awesome

High Performing DevOps teams

More reliable

The 2016 State of DevOps Report

@nicolefv

Page 12: The Data on DevOps: Making the Case for Awesome

IT Performance over the years

The 2016 State of DevOps Report

@nicolefv

Deploy Frequency Change Lead Time Mean Time to Recover

Page 13: The Data on DevOps: Making the Case for Awesome

DevOps promises – and delivers More throughput

More stability

In tandem. Without the tradeoffs that ITIL calls for.

Let’s talk about what this means for us@nicolefv

Page 14: The Data on DevOps: Making the Case for Awesome

High Performing DevOps teams

More agile What does this mean for:200xMore frequent deployments

The 2015 DevOps Survey of Practice and its resulting database are the property of Puppet Labs, Inc. and Gene Kim and Associates, LLC. All rights reserved.

New content deliveryValue/savings around A/B testingValue around speed to marketCompliance / regulatorySecurity2,555x

Faster lead times@nicolefv

Page 15: The Data on DevOps: Making the Case for Awesome

Evaluating well-designed and executed experiments that were designed to improve a key metric, only about 1/3were successful at improving the key metric!

Online Experimentation at Microsoft, Kohavi et al http://stanford.io/130uW6X@nicolefv

Page 16: The Data on DevOps: Making the Case for Awesome

High Performing DevOps teams

More reliable What does this mean for:3xFewer deployfailures

The 2016 State of DevOps Report

Value/savings around reliabilityValue/savings around uptimeComplianceSecurityReputation around uptime,

compliance & security24xFaster MTTR

@nicolefv

Page 17: The Data on DevOps: Making the Case for Awesome

@nicolefv

Page 18: The Data on DevOps: Making the Case for Awesome

"Fixing it as soon as possible or having compensating controls in

place days before could have saved this entire breach from occurring in the first place."

Page 19: The Data on DevOps: Making the Case for Awesome

We know:

• IT Performance is comprised of throughput and stability, and both are possible without tradeoffs

• IT Performance contributes to org performance ($$$)

So:What drives IT and Organizational Performance?

@nicolefv

Page 20: The Data on DevOps: Making the Case for Awesome

IT Performance

Org Performance???

@nicolefv

Page 21: The Data on DevOps: Making the Case for Awesome

IT Performance

Org Performance???

1.

@nicolefv

Page 22: The Data on DevOps: Making the Case for Awesome

IT Performance

Org Performance???

2. 1.

@nicolefv

Page 23: The Data on DevOps: Making the Case for Awesome

IT Performance

Org Performance???

2. $$$1.

@nicolefv

Page 24: The Data on DevOps: Making the Case for Awesome

Continuous Delivery makes our work better…

The 2016 State of DevOps Report

@nicolefv

Page 25: The Data on DevOps: Making the Case for Awesome

“We never had testability before. We have it now. We have this experience and know this stuff is working, and working with controls.”

– Product Owner for Yahoo Chef implementation

Automated configuration and deployment of 250,000 nodes

Can deploy up to 140k node configurations in 8 hours.

Can patch entire infrastructure within 6 hours of a patch being made available

@nicolefv

Page 26: The Data on DevOps: Making the Case for Awesome

Continuous Delivery makes our work better… and makes us feel better!

@nicolefv

https://devops-research.comForsgren, N., J. Humble (2016). "The Role of Continuous Delivery in IT and Organizational Performance." In the Proceedings of the Western Decision Sciences Institute (WDSI) 2016, Las Vegas, NV. Available at SSRN: http://ssrn.com/abstract=2681909

Page 27: The Data on DevOps: Making the Case for Awesome

Microsoft Engineering: DevOps LessonsThiago Almeida -- DevOps Days London, 2016

@nicolefv

Work/Life ScoresBefore CD: 38%After CD: 75%

Source: https://vimeo.com/165184757

Page 28: The Data on DevOps: Making the Case for Awesome

Continuous Delivery makes our work better…

The 2016 State of DevOps Report

@nicolefv

Page 29: The Data on DevOps: Making the Case for Awesome

But what else drives IT Performance?

The 2015 DevOps Survey of Practice and its resulting database are the property of Puppet Labs, Inc. and Gene Kim and Associates, LLC. All rights reserved.

@nicolefv

Page 30: The Data on DevOps: Making the Case for Awesome

Lean Management practices

The 2015 State of DevOps Report

@nicolefv

Page 31: The Data on DevOps: Making the Case for Awesome

“If it moves, graph it.” - Michael Rembetsy, Vice President Operations, Etsy

@nicolefv

Page 32: The Data on DevOps: Making the Case for Awesome

“I was trying to figure out why my team was working themselves to death but not getting anything done… By implementing WIP limits, we were able to focus on our work. Finishing work feels better than sprinting and feeling like a hero in the moment, because that’s only a moment.”

- Julia Wester, Development Manager for Turner Sports, Turner Broadcasting

@nicolefv

Page 33: The Data on DevOps: Making the Case for Awesome

Lean Product Management practices

The 2016 State of DevOps Report

@nicolefv

Page 34: The Data on DevOps: Making the Case for Awesome

Organizational Culture

@nicolefv

Pathological Power-oriented

BureaucraticRule-oriented

GenerativePerformance-oriented

Low cooperation Modest cooperation High cooperation

Messengers shot Messengers neglected Messengers trained

Responsibilities shirked Narrow responsibilities Risks are shared

Bridging discouraged Bridging tolerated Bridging encouraged

Failure leads to scapegoating

Failure leads to justice Failure leads to inquiry

Novelty crushed Novelty leads to problems

Novelty implemented

Page 35: The Data on DevOps: Making the Case for Awesome

• I am glad I chose to work for this organization rather than another company.

• I talk of this organization to my friends as a great company to work for.• I am willing to put in a great deal of effort beyond what is normally

expected to help my organization to be successful.• I find that my values and my organization's values are very similar.• In general, the people employed by my organization are working toward

the same goal.• I feel that my organization cares about me.

Adapted from adapted from Atreyi Kankanhalli, Bernard C.Y. Tan, and Kwok-Kee Wei (2005), “Contributing Knowledge to Electronic Knowledge Repositories: An Empirical Investigation,“ MIS Quarterly, 29, 113-143.

Identity

@nicolefv

Page 36: The Data on DevOps: Making the Case for Awesome

Google Team Performance

Surprise! No magicalformula for what makesthe perfect team

Page 37: The Data on DevOps: Making the Case for Awesome

Intuit

“By installing a rampant innovation culture, we performed 165 experiments in the peak three months of tax season.

Our business result? Conversion rate of the website is up 50%. Employee result? Everyone loves it, because their new ideas can make it to market. ”

- Scott Cook, Intuit founder@nicolefv

Page 38: The Data on DevOps: Making the Case for Awesome

Amazon

“I think building this culture is the key to innovation. Creativity must flow from everywhere. Whether you are a summer intern or the CTO, any good idea must be able to seek an objective test, preferably a test that exposes the idea to real customers. Everyone must be able to experiment, learn, and iterate.”

- Greg Linden

@nicolefv

Page 39: The Data on DevOps: Making the Case for Awesome

IT Does Matter

• Times – and IT – have changed• DevOps is good for Organizations• DevOps is good for IT• And then some detail: What drives this change?

• Technical practices (hint: Continuous Delivery)• Management practices (hint: from Lean Management)• Culture and identity

@nicolefv

Page 40: The Data on DevOps: Making the Case for Awesome

For more information:

@nicolefv

Sign up for our ROI whitepaper & get peer-reviewed researchdevops-research.com

Page 41: The Data on DevOps: Making the Case for Awesome

[email protected]@nicolefv

nicolefv.com

Thank you

Page 42: The Data on DevOps: Making the Case for Awesome

Appendix

Page 43: The Data on DevOps: Making the Case for Awesome

10 deploys per dayDev & ops cooperation at Flickr

John Allspaw & Paul HammondVelocity 2009

That was then…

@nicolefv

Page 44: The Data on DevOps: Making the Case for Awesome

Amazon Deployment Stats (production & host environments only)

This is now…

1,079Max deploysIn a single hour

Every 11.6 seconds!

10,000Mean # hosts receiving Deploys simultaneously

30,000Max # hosts receivingDeploys simultaneously

@nicolefv

Page 45: The Data on DevOps: Making the Case for Awesome

Etsy Code DeploymentWhat once required 6-14 hours and an “Army”

…Now takes 15 minutes and 1 person

This is now…

2013 Mike Brittain, Continuous Deployment: The Dirty Details3/2014 Daniel Schauenberg , Qcon London4/2014 tweet @philkates

30+Deploys per day2013

50Deploys per dayMarch 2014QCon London

80-90Deploys per dayApril 2014Chef Conf

@nicolefv