451 Research: Data Is the Key to Friction in DevOps

© 2017 Delphix. All Rights Reserved. Private and Confidential.© 2018 Delphix. All Rights Reserved. Private and Confidential.
Data Is the Key to Friction in DevOps

• Introduction
• 451 Research Observations and Findings
• Delphix and DataOps
• Choice Hotel’s Journey to remove Data Friction in DevOps
Agenda
3

DevOps means:
• Enterprise priority
• Faster code releases
• Significant code changes
• Expectations of improvement
• Need for more stakeholders - including dbas and data analytics teams
5

Report: Voice of the Enterprise: Cloud Transformation, Vendor Evaluations - Quarterly PowerPoint Charts and Figures
Source: 451 Research, Voice of the Enterprise: Cloud Transformation, Vendor Evaluations 2017
Created for Steven Moretti, The 451 Group from IP 144.121.3.150
© 451 Research, 2000-
2017
Q51. Please describe the level of usage within your organization
for each of the following technologies.

Q15: By what frequency do you deploy code changes or releases?
7
22%
49%
22%
5%
2%
Weeks Days Hours Minutes Seconds

Size of app changes by industry
16%
5%
9%
21% 20%
8%
67%
63%
76%
59%
64%
75%
17%
32%
15%
21%
16% 17%
Total (n=350) High-Tech (n=41) Professional Services
(n=33)
Healthcare (n=29) SaaS (n=25) Trad. Retail (n=24)
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
Large, complex changes
Regular, significant changes
Small, simple changes
Q17. How large/significant are the software application changes that are pushed in your CI/CD workflow?

Q16: What factor of performance improvement in time-to-deploy do you expect
from your CI/CD implementation?
9
8%
29%
36%
15%
7%
5%
2x 3x 4x 5x More than 5x More than 10x

Q25: Beyond IT operations and developers, what stakeholders are involved in your
modern IT management or DevOps initiatives and strategies?
10
64%
46%
66% 66%
1%
0%
10%
20%
30%
40%
50%
60%
70%
Database administrators and
data analytics teams
Lines of business managers Technology center of excellence
or similar group
Security Other (please specify)

CLOUD TRANSFORMATION
WORKLOADS AND KEY PROJECTS 2017
CLOUD TRANSFORMATION
WORKLOADS AND KEY PROJECTS
2017
Q11. Thinking of all the applications your organization runs, what percentage are currently running in the following environments?
Q12. Thinking of all the applications your organization runs, what percentage will be running the following environments two years from now?
Percent of Workloads Running in Cloud
Cloud-Familiar Respondents Using Cloud Computing
11
Source: 451 Research, Voice of the Enterprise: Cloud Transformation, Workloads and Key Projects 2017
45%
24%
42%
28%
58%
38%
62%
43%
55%
37%
64%
49%
6%
7%
13%
14%
11%
15%
11%
14%
15%
17%
15%
16%
7%
12%
8%
9%
5%
6%
6%
8%
6%
10%
6%
9%
6%
11%
20%
27%
14%
21%
14%
20%
8%
10%
6%
11%
36%
46%
18%
23%
11%
20%
7%
15% 17%
27%
10%
16%
SaaS
IaaS
Hosted Private Cloud
On-premises Private Cloud
Non-Cloud
Today In 2 Years Today In 2 Years Today In 2 Years Today In 2 Years Today In 2 Years
n=80 n=80 n=164 n=243
Email and
Collaborative
Web and
Media
Today In 2 Years
Data and
Analytics
Application
Development
Business
Applications
Shared IT
Workloads
n=118 n=122 n=163 n=74 n=73 n=248 n=375 n=375

Delphix
12

Massive,
Sprawling Datasets
Increasing
Governance & Risk
Enterprise Silos &
Manual Process
Data-driven
Decisions
Faster Dev &
Deployment
Increased Cloud
Adoption
13

© 2017 Delphix. All Rights Reserved. Private and Confidential. 14
We give enterprises the ability to deliver
high quality software at the pace needed
to thrive in the Digital Economy.
We free builders from the constraints of limited data
environments, empowering them with self-service
access to real data without jeopardizing security.

Critical User Needs in the Modern Enterprise
15
• Protect Sensitive Data
• Governance & Compliance
• Delivery Automation
• Manage Data Sprawl
• Reduce Storage Overhead
DATA CONSUMERS
• Instant Data Access
• Personal Environments
• Self-Service Controls
• Full, Realistic Data
• Process & Tool Aligned
DATA PROVIDERS

Delphix Dynamic Data Platform
Fast. Secure. Everywhere.
16

Production
Development
Manual
periodic
refresh…
… with test
data added
to SIT
v79
v79
v79
Git Repo
(code + schema changes)
+100 rows
+ v80
v80
v80
v80
Jenkins / build runs
schema migration
and unit tests
schema migration
and functional tests,
which require test data
Destructive testing
slowly mutates test
database
Production deploy
Testing

… time passes,
production data changes
and development
and testing continue

Production
Testing
Development
Another
manual
refresh…
… with test
data added
to SIT
v87
v87
v87
Git Repo
+100 rows
+ v88
schema migration
and unit tests
schema migration
and functional tests,
which require test data
Destructive testing
slowly mutates test
database
v88
v88
v88
Multiple rounds of
testing has caused
heavy data mutation,
putting quality at risk
Production deploy

Production
Development
DBA sets up
Delphix one time,
prod now continuously
syncs and is masked v79
v79
v79
Git Repo
+100 rows
+ v80v80
v80
v80
Jenkins / build
refreshes database
using Delphix APIs…
refreshes database
Production deploy
… then runs
schema migration
and unit tests
… runs schema migration
… inserts test data
… and runs functional tests
Testing

Production
Development
v87
v87
v87
Git Repo
+100 rows
+ v88
v88
Jenkins / build
refreshes database
refreshes database
Production deploy
… then runs
schema migration
and unit tests
… runs schema migration
… inserts test data
… and runs functional tests
v88
v88
Development and testing
databases have been kept
fresh during iterative
development
Testing

Data Friction
in DevOps
Page 25

Introduction
Principal Systems Engineer
Choice Hotels
Page 26

Manual steps to configure physical and virtual infrastructure
Before 2017
Page 28
VM

Choice Private Cloud Introduced in 2014
• Good tool for infrastructure
• 98% VMs created with this tool, steady increase year over year
Before 2017
Page 29

Before 2017
Page 30
Task automation, increase
the delivery speed of
components
VM

• Focus on Programmable Infrastructure and Platform
– API enabled and Leveraging Jenkins
• vRA
• Load Balancers
• Firewalls (Initial Stages)
• Storage
• Delphix virtual DB provisioning and management
• Increased Programmable Middleware deployment and operation
2017-2018
Page 31

2017-2018
Page 32
Approved
Scripts &
Pipeline
Rest API
vRA [vRealize Automation]
vRO [vRealize Orchestration]
vCENTER
VM VM VM VM VM
IP Address / DNS
Infrastructure
Management
ADC
GREEN = Deployed and functioning
ORANGE = Research, Validation, or Configuration Required

• Our ultimate goal is to offer ‘IT as a service’ to our Build and Run teams,
across the layers of our environments
Where Delphix fits in our Environment Strategy
Page 33
aPaaS
Application Platform
as a Service
PaaS
Platform as a Service
IaaS
Infrastructure as a
Service
• Ability spin up an application in a dev/test environment, through automation
• Jenkins for continuous integration
• Git for code repository
• Maven for build automation
• …etc.
• Ability to spin up middleware and databases in a dev/test environment,
through automation
• vSphere for…
• Python
• Delphix
• … etc.
• Ability to spin up compute, storage and network in a dev/test environment,
through automation
• vSphere
• Puppet/SaltStack
• Vagrant
• …etc.

Data as a Service (DaaS)
Page 34 Delphix Company Confidential November 2016
Automation
Configuration
Machine
Virtualization
Data
Management

What are we solving with Delphix?
• Choice struggles with the ability to manage test environments & data refreshes
• Test data should mirror production data with masked PII & PCI information
• Build Leaders would like the ability to access multiple test environments for parallel testing
• Environments require multiple copies of the same data, causing over provisioning of
resources
• Full backups of critical databases can take up to 6 hours and recovery can take up to 9 hours,
scripts to mask/cleans data
• Backup environment is out of capacity

Solutions to Issues
• Pull top 10 databases (CRS, CIS, cA, SmartRates, etc.) into Delphix appliances
• Make virtual copies for each tier (DEV, QA, LT, etc.)
• Continue to backup production databases with Veritas; leverage Delphix for lower tier
backups
• Implement data refresh process at the beginning of each cycle for more consistent results in
testing
• Implement more robust data masking/cleansing process
• Teach developers how to rollback changes, reducing dependencies on database administrators

Delphix at Choice
Masking Engine
Source Databases
Delphix Engine
MSSQL Pre-Provisioning
Target Server *
Delphix Engine
PCI Internal DMZ
REG Internal DMZ
DEV/QA/LT DMZ
MSSQL
Informix
Oracle
Target Databases
Source Databases
MSSQL
Informix
Oracle
MSSQL
Informix
Oracle
Replication to
Non-PCI zone
engines

Thank You
38

451 Research: Data Is the Key to Friction in DevOps

Recommended

Recommended

More Related Content

What's hot

What's hot (20)

Similar to 451 Research: Data Is the Key to Friction in DevOps

Similar to 451 Research: Data Is the Key to Friction in DevOps (20)

More from Delphix

More from Delphix (20)

Recently uploaded

Recently uploaded (20)

451 Research: Data Is the Key to Friction in DevOps