Big Data – A New Testing Challenge

@esconfs|#esconfs

Adam KnightTeradata [email protected]

Big Data – A New Challenge for Testing

@esconfs|#esconfs

Cloud Testing

Web Testing

Mobile App Testing

Big Data Testing

Big Data – a Niche Market?

@esconfs|#esconfs

Cloud Testing

Web Testing

Mobile App Testing

Big Data Testing


@esconfs|#esconfs


Demand for application development skills in big data positions increased 396% in the first 6 months of 2015 as compared to the previous year

Source: Wanted analytics/Forbes – figures based on jobs quoting ‘Big Data’

@esconfs|#esconfs

Big Data is Coming

The Internet of Things is the term used to describe network connectivity and communication between machines and devices

@esconfs|#esconfs

Big Data is Coming

Gartner, Inc. forecasts that 4.9 billion connected things will be in use in 2015, up 30 percent from 2014, and will reach 25 billion by 2020 Source: Gartner (November 2014)

These are all connecting to share data

@esconfs|#esconfs

Big Data is Coming

There is an expectation in modern devices and systems of :-• Connectivity with other systems

• Data collection• Exposure to Analytic capability

Business AssetInsight Business

Data

@esconfs|#esconfs

Big Data is ComingConnected

devices

Activity data capture

Data collecting monitors

Storage Technology

Analytical Technology

Management and Control

Systems

Inevitably we’ll see an increase in….

…and the need to test them.

@esconfs|#esconfs

So what makes it Big Data?

Data that is “Too Big”

…but “Too Big” for what?

@esconfs|#esconfs

Too Big to…..?

Back Up?

Manipulate?

Comprehend?

Count the Individual?No… but too big to test every individual

No… but too big to manipulate the whole dataset

No… but too big to duplicate production for testing

No… but too big to consider scale at all levels at once

@esconfs|#esconfs

Too Big to Test…..?

No… But we need a Change of Perspective…

@esconfs|#esconfs

Scalability LayersFrom To

Records Partitions

Data Metadata

Machines Clusters

@esconfs|#esconfs

Scalability Layers

@esconfs|#esconfs

Non-Functional?What we may previously have described as ‘non-Functional’ characteristics are the cornerstone of Big Data technologies

• Scalability is a function of a Big Data system

• Performance is a function of a Big Data system

@esconfs|#esconfs

Non-Functional?User stories/requirements are often focussed around “non-functional” characteristics

“As a query user I want my lookup queries to run in 50% of the time as they do at present.”

“As a cluster administrator I want to store over 1 million data files with no loss of import performance”

We can’t leave testing of performance and scalability until ‘later’. Our everyday testing must focus on these.

@esconfs|#esconfs

Capturing/Checking

Capturing

@esconfs|#esconfs

Capturing/Checking

CapturingChecking

@esconfs|#esconfs

ParalleliseRandomiseIterate

Automation Approaches

@esconfs|#esconfs

ParalleliseRandomiseIterateParameterise

Parameterized Test Output


@esconfs|#esconfs



@esconfs|#esconfs



@esconfs|#esconfs

Hadoop

Cutting edge Hard to get experience

Evolving Volatility

Varied Interfaces Complexity

Open Source/Multiple Vendors Inconsistency

@esconfs|#esconfs

Thank You

Adam KnightTeradata [email protected]

Big Data – A New Challenge for Testing

@esconfs|#esconfs

ImagesFractal Mountain http://www.miqel.com/fractals_math_patterns/visual-math-natural-fractals.htmlEverest - http://en.wikipedia.org/wiki/File:Everest_kalapatthar_crop.jpg Too Big to Count - http://commons.wikimedia.org/wiki/File:Colony_of_aptenodytes_patagonicus.jpg Too Big Spoon - http://www.flickr.com/photos/oskay/426004828 ; Too Big to Back Up - http://ratien.livejournal.com/205294.html ; Quasar: https://en.wikipedia.org/wiki/Quasar#/media/File:Artist%27s_rendering_ULAS_J1120%2B0641.jpgTrees: https://www.flickr.com/photos/robwatling/3437773221 https://www.flickr.com/photos/brian-m/167546353/ https://www.flickr.com/photos/freefoto/728649489 acorn https://www.flickr.com/photos/randihausken/2881856339/ http://www.state-journal.com/6043637966 Teracotta army: https://www.flickr.com/photos/3708564147 https://www.flickr.com/photos/11523510726 https://www.flickr.com/photos/11523510726https://www.flickr.com/photos/4550658574 Whurlitzer - http://www.flickr.com/photos/27485954@N07/5431975868 : ZoeTrope - http://en.wikipedia.org/wiki/File:Zoetrope.jpgRoulette - http://middlebororemembers.blogspot.co.uk/2013/02/banks-kicking-gambling-addiction.htmlStopwatch - http://en.wikipedia.org/wiki/File:Stopwatch2.jpg GPS - https://www.flickr.com/photos/vikapproved/13950372514/Seeds - http://www.flickr.com/photos/yimhafiz/5035250840/ ; Segments - http://www.flickr.com/photos/beta-j/4432637412Hadoop Elephant: Publicly availableRest : Me

Big Data – A New Testing Challenge

Software