This document summarizes Karen Lopez's presentation on being a lazy data modeler. The presentation argues that automating repetitive tasks through tools' automation features and scripting languages like PowerShell allows data modelers to spend more time on higher-level work. Lopez provides 10 tips for data modelers to identify mindless tasks for automation, learn automation skills, and implement automation iteratively. The goal of laziness is presented as producing higher-quality models and databases through focusing one's work, rather than doing menial tasks manually.
2. Karen Lopez
@DATACHICK
May 2015
www.dataversity.net
www.datamodel.com 2
Karen López
Karen has 20+ years of data and information architecture
experience on large, multi-project programs.
She is a frequent speaker on data modeling, data-driven
methodologies and pattern data models.
She wants you to love your data…
She is very, very lazy
How Lazy Are You?
...so let’s get to know you….
3. Karen Lopez
@DATACHICK
May 2015
www.dataversity.net
www.datamodel.com 3
Attendees, be part of the webinar
Use Q&A
for formal
questions
Use chat
to discuss
with each
other
Plan for Today
Why topic?Why topic?
What? Lazy? What the Heck?What? Lazy? What the Heck?
Some Demos, Screenshots & What NotSome Demos, Screenshots & What Not
10 Tips for Being More Lazy10 Tips for Being More Lazy
5. Karen Lopez
@DATACHICK
May 2015
www.dataversity.net
www.datamodel.com 5
But it’s not about free time
•Forensics
•Serving our “customers”
•Better quality data models
•Better databases
•Providing better support to teams
•Making models more accessible
•Removing obstacles to data model use
•Doing mindful tasks and activities
It’s about better modeling time…more time for
6. Karen Lopez
@DATACHICK
May 2015
www.dataversity.net
www.datamodel.com 6
Lazy Data Modeler = Better Data Modeler
Still work
hard On more
important tasks
Why data modelers don’t want to automate
There’s a
learning curve
No one shares
their scripts
“I’m not a
programmer”
They don’t
know they
can
7. Karen Lopez
@DATACHICK
May 2015
www.dataversity.net
www.datamodel.com 7
They don’t know they can
Have never
clicked on
that feature
Were
perplexed
when they
did
No idea what
to do when
they click
Tried it once,
broke their
model and
never want
to come back
Tried to do it,
it was a huge
timesuck, so
they gave up
There’s a learning curve..
Why,
yes,
there
is…
Start with samples and shared scripts
Do a “Hello World!”
Spend 20 minutes a day or week learning a bit more
Or spend 20 minutes day or week making a business case for developer
support
Get some training
Join an online community/forum
10. Karen Lopez
@DATACHICK
May 2015
www.dataversity.net
www.datamodel.com 10
Question Break
IF Question(You) THEN Answers(Datachick)
END IF
“I’m not a programmer”
Great! Not a problem!
Some tools…
require real
application
development
skills.
require
scripting level
skills
will record
your
keystrokes and
generate a
script – Excel,
for instance.
come with
sample
macros/scripts
provide places
for
organization
share their
macros and
scripts
11. Karen Lopez
@DATACHICK
May 2015
www.dataversity.net
www.datamodel.com 11
No one shares their scripts
History of non-sharing
Online Communities
Github, etc.
It’s time to join this century.
Yes, there are legal issues. But these macros aren’t
any more sensitive / proprietary than other scripts
that are shared widely.
Kinds of Lazy
Internal Model CRUD
Naming
Creating columns
Applying Indexes, Constraints
…more
External productivity
Printing
Generating Reports
Generating Images
Making Backups
Managing files, templates, config files, etc.
12. Karen Lopez
@DATACHICK
May 2015
www.dataversity.net
www.datamodel.com 12
Automated Naming standards
Let the computer apply your crazy meta data stuffing schemes
http://www.datamodel.com/index.php/2012/10/10/metadata-
stuffing-why-i-hate-tbl_-for-table-names/
Deal with physical constraints of your DBMS
Case, spaces, special characters, length, etc.
All the tools have something that does this, and they are similar.
But sometimes the naming utilities aren’t enough (more later)
Let’s look at some tools…
13. Karen Lopez
@DATACHICK
May 2015
www.dataversity.net
www.datamodel.com 13
CA ERwin Data Modeler
Full blown API
Active Scripting
Visual Basic for Applications
Object oriented application features and requirements
Documented online ERwin API Reference Guide support.ca.com
Erwin-knowledgebase.com
14. Karen Lopez
@DATACHICK
May 2015
www.dataversity.net
www.datamodel.com 14
Automated Naming standards
Let the computer apply your crazy meta data stuffing schemes
http://www.datamodel.com/index.php/2012/10/10/metadata-
stuffing-why-i-hate-tbl_-for-table-names/
Deal with physical constraints of your DBMS
Case, spaces, special characters, length, etc.
All the tools have something that does this, and they are similar.
But sometimes the naming utilities aren’t enough (more later)
SAP PowerDesigner
Java, VBScript, C, other languages
Executed inside the tool
Text files that can be edited outside the tool
Documentation on the web infocenter.Sybase.com
16. Karen Lopez
@DATACHICK
May 2015
www.dataversity.net
www.datamodel.com 16
Question Break
IF Question(You) THEN Answers(Datachick)
END IF
So why Lazy? Mindless tasks take up a lot of
time
You were hired for your brain,
not your good looks
More time for modeling, not
printing, reporting, etc.
More time to help devs &
DBAs
17. Karen Lopez
@DATACHICK
May 2015
www.dataversity.net
www.datamodel.com 17
PowerShell is your Lazy Enabler
• Windows Feature
• Automates EVERYTHING
• Creating VMs and configuring
them
• Running your data modeling tool
macros while you are sleeping
• Backing up files, databases, etc.
• Just about anything in Windows
and Azure and….
Karen’s Rules for Being Lazy
Don’t spend time doing things that a computer is faster
and better at
Automation is your friend
Don’t try to automate everything at once
Don’t try to rebuild an entire data modeling tool in a script
Focus mindful things, not mindless ones
If you’ve automated it, you must ask vendors to make it a
feature in their tool
18. Karen Lopez
@DATACHICK
May 2015
www.dataversity.net
www.datamodel.com 18
So let’s summarize:
• Automating boring tasks makes you happier.
• Happier Data Architects are better Data Architects
• Automated recurring, boring tasks make bosses
happier
• Automating tasks makes for more accurate work
• Saving time for you and your team members
makes everyone happier.
10 Tips for Being a Lazy Data Modeler
1. Learn automation features in your tools
2. Use automation features in your tools
3. Learn PowerShell
4. Never run a script on your production models
without testing and understanding it completely
5. Ask for developer support
19. Karen Lopez
@DATACHICK
May 2015
www.dataversity.net
www.datamodel.com 19
10 Tips for Data Modelers
6. Examine your activities. Question all of them.
7. Find mindless tasks that are your TOP candidates for
automation
8. Free up junk modeling time to allow you to do mindful
work
9. Think in terms of iterative, incremental improvement,
not big bang automate the whole world first
10.Be lazy. All the time. Every day. Get more lazy.
More Resources
The Best DBA is a Lazy DBA: Guide to the Minimalist DBA
(with Thomas LaRock)
http://fundamentals.sqlpass.org/MeetingDetails.aspx?EventID=853
PowerShell
https://technet.microsoft.com/en-us/scriptcenter/powershell.aspx
Github
https://github.com/
20. Karen Lopez
@DATACHICK
May 2015
www.dataversity.net
www.datamodel.com 20
Call to Action!
1. Find 3 activities you do now that are mindless junk
modeling
2. Search for scripts/macros that automate them
• Web Search
• Ask on forums
• Find a similar one
3. Make it your own
4. Use it
Embarcadero T-shirt Draw