guide to the java ecosystem

7/24/2019 Guide to the java ecosystem

1/37

THE DZONE GUIDE TO

2015 EDITION

THE JAVA

ECOSYSTEM

BROUGHT TO YOU IN PARTNERSHIP WITH


2/372

ZONE.COM/GUIDES DZONES 2015 GUIDE TO THE JAVA ECOSY

DZONES 2015 GUIDE TO THE JAVA ECOSYSTEM

TABLE OF CONTENTS

20 yearsa lieime or a programming language, never

mind a whole platorm. This year, we celebrate Javas

20th birthday, and what better way to celebrate than to

release DZonesfirstGuide to the Java Ecosystem.

I can sill remember the 10THJavaOne. There was

energy and a belie that Java would ulfill its mission

to give developers a true chance to Write Once,

Run Anywhere. All this while Sun Microsystems

struggled to find a place in the market amid a changing

economy and a rapidly evolving Internet. Today, Java is

everywhere, and it is the workhorse o enterprises and

independent developers alike. It is open. It is more than

just a language. And ater all this ime, Java remains the

most popular language or developers (Tiobe) and one o

the most acive or job hunters (Indeed.com).

Long beore DZone.com, we started Javalobby.org: an

independent group o Java developers eager to talk

about this new language that didnt require memory

management and had the backing o some powerul

organizaions. Weve evolved our vision o Javalobby

over the years (now our Java topic portal), but it remains

one o the largest secions o our site and one o the

primary desinaions or Java developers on the web. Its

because o Javas role in the very oundaion o DZone

that finding the right itle and ocus or this guide was

one o the hardest decisions our editorial team has had

to make this year.

This guide, which youre hopeully enjoying today,

ocuses on trying to understand the trends about how

developers such as yoursel are using the Java platorm.

What version o the platorm are you using? Are you

using a ull JEE stack or just using simpler tools like

Spring? What percentage o developers preer Tomcat

over Websphere? Staisics like thesealong with

expert analysis and aricles rom industry leaders

await you in the 2015 Guide to the Java Ecosystem.

Thanks, as always, or your support, and pleaseconinue to visit DZone.comdaily.

P.S. Were always hiringget in touch i youd like to

help build the platorm that powers DZone.com and

communiies or LinkedIn, Microsot, and a variety o

other organizaions.

MATT SCHMIDTPRESIDENT AND CTO

[email protected]

EXECUTIVE SUMMARY

KEY RESEARCH FINDINGS

WHY JAVA 8?BY TRISHA GEE

DIVING DEEPER INTO THE JAVA ECOSYSTEM

FIRST STEPS IN JAVA MICROSERVICESBY IVAR GRIMSTAD

PRODUCTION DEBUGGING IS NOT A CRIMEBY ALEX ZHITNITSKY

JAVA POPULARITY: BY THE NUMBERSINFOGRAPHIC

REACTIVE TRENDS ON THE JVMBY JONAS BONR

JAVA ECOSYSTEM EXECUTIVE INSIGHTSBY TOM SMITH

THE POWER, PATTERNS, AND PAINS OF MICROSERVICESBY JOSH LONG

JAVA BEST PRACTICES CHECKLIST

SOLUTIONS DIRECTORY

DIVING DEEPER INTO FEATURED JAVA ECOSYSTEMSOLUTIONS

GLOSSARY

3

4

6

9

12

16

18

20

24

28

32

33

36

37

EDITORIALJohn [email protected]

G. Ryan SpainDIRECTOR OF PUBLICATIONS

Mitch PronschinskeSR. RESEARCH ANALYST

Matt WernerMARKET RESEARCHER

Moe LongMARKET RESEARCHER

John WalterEDITOR

Allen CoinEDITOR

Tom SmithRESEARCH ANALYST

BUSINESSRick RossCEO

Matt SchmidtPRESIDENT & CTO

Kellet AtkinsonVP & PUBL ISH ER

Matt OBrianDIRECTOR OF BUSINESS

DEVELOPMENT

Jane ForemanVP OF MARKET ING

Alex [email protected] OF MAJOR

ACC OUN TS

Chelsea BosworthMARKETING ASSOCIATE

Chris SmithPRODUCTION ADVISOR

Jim HowardSALES ASSOCIATE

Chris Brumfeld

CUSTOMER SUCCESS

ASSOC IATE

ARTAshley SlateDESIGN DIRECTOR

Yassee MohebbiGRAPHIC DESIGNER

Special thanks to our texperts Paul Bakker, Reza

Rahman, Markus Eisele, A

Gupta, Josh Long, Ivan StIvanov, Marcus Lagergren

and our trusted DZone Mo

Valuable Bloggers for all t

help and feedback in mak

this report a great succes

WANT YOUR SOLUTION TO BE FEATURED IN COMING GUIDES?Please [email protected] submission information.

LIKE TO CONTRIBUTE CONTENT TO COMING GUIDES?Please contact [email protected] consideration.

INTERESTEDIN BECOMING ADZONE RESEARCH PARTNER?Please contact [email protected] information.

Dear Reader,
http://www.dzone.com/guideshttp://dzone.com/http://dzone.com/mailto:?subject=research%40dzone.commailto:research%40dzone.com?subject=mailto:sales%40dzone.com?subject=mailto:research%40dzone.com?subject=mailto:research%40dzone.com?subject=mailto:research%40dzone.com?subject=mailto:?subject=sales%40dzone.commailto:?subject=sales%40dzone.commailto:?subject=sales%40dzone.commailto:?subject=sales%40dzone.commailto:?subject=sales%40dzone.commailto:?subject=sales%40dzone.commailto:?subject=research%40dzone.commailto:[email protected]:research%40dzone.com?subject=mailto:research%40dzone.com?subject=mailto:sales%40dzone.com?subject=mailto:research%40dzone.com?subject=http://dzone.com/http://dzone.com/http://www.dzone.com/guides


3/37DZONES 2015 GUIDE TO THE JAVA ECOSYSTEM


DZone surveyed more than 600 IT proessionals and

produced several expert aricles or our 2015 Guide to

the Java Ecosystem to give Java organizaions greater

insight into the eects o major trends on the larger Java

development community and to discover common patterns

and technology stacks currently in use across the Java

industry. In this summary you will learn about the current

state o the Java platorm and the ecosystem o tooling

around the Java language.

RESEARCH TAKEAWAYS

01 DEVELOPERS ARE MOVING TO NEW VERSIONS OF

JAVA MORE QUICKLY

Data: 53% o respondents say they their new apps will useJava 8 in the next six months and 35% are going to startconvering exising apps to Java 8 in that same ime rame. 62%o respondents said their exising apps are in Java 7 and 20%already have exising apps in Java 8.

Implicaions: A slight majority o Java organizaions seem to be

moving to the latest version o Java or new applicaions. Thisshows a bit more orward movement in the Java community thanthere has been in years past. Large swaths o Java companieshave sizable legacy codebases due to their organizaions ageand the constant progression o the language, which has letsome o them behind. There will always be a large number oorganizaions where there isnt enough risk or enough resourcesto moivate an update to their legacy code, but its surprising tosee significant numbers o Java companies moving their apps tothe newest version.

Recommendaions: The act that Oracle has declared Java SE7 end-o-lie is a significant contributor to this trend towardsaster migraion. To maintain a high level o security and takeadvantage o new language eatures that other programminglanguages have had or a long ime, migraion to the latest versionJava is always the best opion i an organizaion can aord it.Trisha Gees aricle in this guide, Why Java 8? illustrates theposiive impact Java 8 can have on your applicaions. Somekey things enicing developers toward Java 8 are uncionalprogramming eatures and perormance improvements.

02 INTEREST IN CONTAINERS & MICROSERVICES IS

GROWING IN THE JAVA WORLD

Data: While nearly hal o respondents havent researchedcontainers or microservices enough to comment on them (43%

or microservices, 46% or containers), 21% want to implement

containers (12% already have) and 15% want to implement

microservices (10% already have). There are more respondents

interested in containersor actually using themthan

those that arent (12% not interested) and the same goes or

microservices (20% not interested).

Implicaions: These trends have made a significant impact in

the Java community over the last 2-3 years, but it will sill take

several more years or containers or microservices to obtain

mainstream adopion in most o the organizaions where they

can be beneficial. However, many o the most successul web

properies menioned previously (Google, Net lix, Twitter, etc.)

have already implemented containers and microservices where

appropriate. These companies are driving trends by sharing

some o their implementaions.

Recommendaions: Developers, architects, and business

execuives need to start researching the potenial benefits that

containers and microservices bring to their applicaions. Some

may decide that they are not appropriate or mature enough or

their systems, but more definiive research must be done so that

a decision can be made. Several technologies have already moved

to support microservicessuch as Spring Boot, Wildly Swarm,and Akka. Docker already has a dominant posiion in the

containerizaion space. A good place to start researching Docker

is by reading Docker and Kubernetes in the Java Ecosystem by

Paul Bakker on DZone. For building microservices in the Java

ecosystem, you should read Ivar Grimstads aricle in this guide,

Building Microservices With Java.

03 JAVA IS ONE OF THE MOST POPULAR PROGRAMMING

LANGUAGES

Data: There are several indexes in tech media that use various

metrics to measure programming language popularity, including

TIOBE, RedMonk, and IEEE. Java is ranked #1 or #2 in all othese major indexes. No other language has rankings that are as

consistently high as Javas across all these indices. On the TIOBE

index, one o the longest running indices, Java has been the number

one language or 11 o the last 14 years, and it is currently number

one by a significant margin. Only 3% o respondents in our Java

ecosystem survey said they are pessimisic about the uture o Java.

Recommendaions: Java has reached a point where its

popularity results in urther popularity. More developers are

using it, more open-source uiliies exist around it, and there

is more maturity and reliability around the platorm and its

ecosystem o tools. As a result, even more organizaions and

developers choose to build applicaions in Java.

Recommendaions: While many thought Java would decline ater

the announcement o Oracles acquisiion o Sun Microsystems

in 2009, it only took a ew years to reestablish itsel. Now with the

release o Java 8, Java is finally catching up with the expressiveness

o C# and now has the ability to be written with less verbosity than

ever beore. People are now opimisic about the uture o Java, and

its being used in many o the most well-respected tech companies.

Its popularity in the sotware industry is expected to coninue

and even grow urther due to its advantages in data analysis and

embedded systems, which will likely make Java a significant

programming language in the Internet o Things industry.

ExecutiveSummary
http://www.dzone.com/guideshttp://www.dzone.com/guides


4/374



01. MOST DEVELOPERS ARE KEEPING UP FAIRLY

WELL WITH NEW JAVA VERSIONSA majority o respondents (58%) said their organizaion was

using Java 8the latest version o Javain at least some

o their new applicaions. 48% said some o their new apps

would be built on Java 7. Only 10% o respondents indicated

the use o Java 6 or below or any new apps, and a minuscule

1% said new apps would use Java 5 or lower. As ar as exising

apps are concerned, Java 8 has not caught up to Java 7; most

organizaions (62%) sill have apps built on Java 7, and almost

hal (47%) have apps in Java 6. Only 20% o respondents said

they had any exising apps built on Java 8. While this data

shows that many developers are eager to move orward

with Java as it evolves, they are hesitant to reactor exising

applicaions already built on a previous platorm. 18% o Java 7

users who have not moved to Java 8 have no plans to start usingJava 8 in the next 6 months, or either new or exising apps.

02. JAVA EE AND SPRING USAGE IS ALMOST EQUALOne o the significant rits in the Enterprise Java community

is the dierence between organizaions that use Spring and

those that use Java EE. Our survey shows that the usage o some

components o the two platorms is pretty close, with Java EE

having a slight edge. Overall, 58% use some version o Spring

and 67% use some version o Java EE. Very ew respondents used

Java EE 5 or lower, or Spring 2.x or lower. The breakdown o each

versions usage can be ound in the charts, but whats interesing

is that almost one-third (32%) o respondents use both.

03. JVM LANGUAGES HAVE A STRONG SHARE OF JAVA

DEVELOPERS57% o respondents have written code in a non-Java, JVM-

based language. Groovy (40% o al l respondents) is the

language that most have tried, with Scala (31%) close behind.

The next highest languages were JRuby and Clojure, which

were both around 6%. For each o the groups who had

tried a specific JVM language, over 80% said they enjoyed

using it. Only Groovy and Scala had large enough groups o

users to make a staisically significant observaion about

user opinions. When asked i they would li ke to use these

languages at their job, 45% o Groovy users said they would,and 55% o Scala users said they would.

04. CERTAIN TOOLS DOMINATE IN MOST AREAS OF

THE JAVA ECOSYSTEMWe asked respondents what their primaryIDE, build tools,and CI engines were. For IDEs, Eclipse is sill king with 54% orespondents using it. 30% use IntelliJ IDEA (most respondents[23%] actually use the paid version, with only 7% using thecommunity ediion). 13% use NetBeans. For build tools, Maven

KeyResearchFindings

02. WHICH OF THE FOLLOWING JAVA PLATFORMS DO YOU USE?01.WHAT VERSIONS OF JAVA ARE BEING USED AT YOUR ORGANIZATION?

More than 600 IT professionals respondedto DZones 2015 Java Ecosystem Survey.Here are the demographics for this survey:

Developers (42%) and Development Leads (25%) were

the most common roles.

63% of respondents come from large organizations

(100 or more employees) and 37% come from small

organizations (under 100 employees).

The majority of respondents are headquartered in

Europe (44%) or the US (32%).

Over half of the respondents (61%) have over 10 years

of experience as IT professionals.

We only surveyed developers who use Java in their

organization. Other language ecosystems for these

organizations include C# (31%), C/C++ (29%),

Universal JavaScript (28%), PHP (26%), Python(26%), and Groovy (23%).

Java 5 or lower

Java 6

Java 7

Java 8

VERSION

1%

10%

48%

58%

11%

47%

62%

20%

FOR NEW APPS FOR EXISTING APPS

JAVA EE 6

JAVA EE 7

SPRING 3.X

SPRING 4.X

39%

36%

38%

34%

JAVA EE SPRING32%




(66%) is much more popular than other compeitors like Ant(18%) and Gradle (11%). Jenkins absolutely dominates the CIspace, with 62% using it. The next highest CI server has lessthan 10% share o respondents. 20% dont use a CI server. Wealso asked respondents to list all t he Java persistence toolsthey use and ound that standard JPA/Hibernate is the mostcommon persistence soluion or Java developers (64%), whilea ew others preer JDBCeither standard (38%) or SpringJdbcTemplate (24%) versions.

05. JSF AND SPRING MVC ARE NECK-AND-NECK IN

THE JAVA WEB FRAMEWORKS RACEThe Java web ramework battle is much closer than thepreviously menioned tooling compeiions. In an August 2015poll, DZone surveyed over 1,300 developers, and the resultsshowed Spring MVC (34.2%) and Java Server Faces (34.5%) ina virtual ie. However, more Java developers are looking orront-end development tools outside o the Java ecosystem.When asked in the 2015 Java Ecosystem Surveyabout the toolsthey use or applicaion ront-ends, respondents actuallychose AngularJS (43%) more than Spring MVC (34%) and JSF(30%). All o Angulars direct compeitorsReact.js, Ember.js,Backbone.jswere under 10%.

06. FULL JEE APP SERVERS ARE NOT COMMONLightweight applicaion servers like Tomcat (68%) and Jetty(27%) have been dominant in Java development or someime. For several years, the trend in Java development hasbeen moving toward the use o only the components that youneed, and JEE eatures in applicaion servers are no excepion[1]. We also ound that most o these app servers are used indevelopment more oten than they are used in producion.Only WebLogic and WebSphere were deployed in producionmore oten than development.

07.JAVA DEVS SLOWLY ADOPTING CLOUD, BUTMICROSERVICES AND CONTAINERS ARE FURTHER OFF

The adopion o cloud inrastructure is sil l modest in Javaorganizaions, with only 37% r unning some percentage o

their applicaions on the cloud during the development phase.

43% ran some percentage o their tests/QA on the cloud, and

42% are using some cloud inrastructure in their producion

environment. When we asked about more recent development

trends like microser vices and containers, only 10% were

using microservices and only 12% were using containers

(specifically, 9% were using Docker). Most respondents haven

researched these two trends enough to even think about

using them (43% or microservices, 46% or containers), and

some dont see a use or them currently or or the oreseeableuture (20% or microservices, 12% or containers). However,

the interest in these two trends is significant with 15%

waning to implement microservices and 21% waning to

implement Docker containers.

08. NEWER SERVER TECHNOLOGIES ARE STILL

BLEEDING-EDGE AND RAREWe asked respondents about the server technologies they

were using, and while most were sil l using Java EE or Spring

primarily, we ound small groups using some o the newer

server technologies in the Java space. 11% said they are using

a containerless method with Spring Boot and Dropwizard.

However, Spring Boot and Dropwizard are used or more

than just containerless deployment, so its definitely not an

indicator o the two products overall popularity. 9% said they

are using Scala and Akka, and 2% are using Vert.x.

[1] bit.ly/1NYf7v

04. WHICH APP SERVER(S) DO YOU USE?

05. WHICH OF THE FOLLOWING SERVER TECHNOLOGIES DO YOU USE?

03. WHAT FRAMEWORKS DO YOU USE FOR WEB APPLICATIONS?

TOMCAT

WILDFLY/JBOSS

JETTY

WEBLOGIC

GLASSFISH/PAYARA

WEBSPHERE

NONE

TOMEE

68%

35%

27%

18%

15%

15%

6%

5%

use a containerless method

with Spring Boot/Dropwizard

useScala/Akka

use Vert.x

STRUTS

JSF

SPRING MVC

SERVER SIDE

CLIENT SIDE

ANGULARJS

BACKBONEJS EMBERJSREACTJS

43%

34%

30%

16%

8%8% 6%
http://www.dzone.com/guideshttp://bit.ly/1NYfj7vhttp://bit.ly/1NYfj7vhttp://bit.ly/1NYfj7vhttp://www.dzone.com/guides


6/376



WhyJava 8?

BY TRISHA GEE

Java 8 came out early last yearand

Java 7 is now end of lifemaking Java

8 the only Oracle-supported opion unil

Java 9 comes out at the end of next year.

However, since organizaions value

stability over trendiness, many of us aresill working with Java 7, or even 6.

Lets look at some features of Java 8, and

provide some arguments to persuade your

organizaion to upgrade.

ITS FASTER

Heres a selling point that might please your boss, the

business, or the operaions guys: youll probably findJava 8 runs your applicaion aster. Generally speaking,

applicaions that have moved to Java 8 see some sort o speed

improvement without any specific work or tuning. This may

not apply to an applicaion that has been highly tuned to a

specific JVM, but there are a number o reasons why Java 8

perorms better:

Performance Improvements in Common Data Structures:

Benchmarks o the ever-popular HashMapshow that

perormance is better in Java 8. These sorts o improvements

are very compellingyou dont need to learn the new

Streams API or lambda syntax or even change your exising

code to get speed improvements in your applicaion.

Garbage Collector Improvements: Oten Java Perormance

is synonymous with Garbage Collecion, and it is certainly

true that poor garbage collecion perormance will impact an

applicaions perormance. Java 8 has substanial changes to

GC that improve perormance and simpliy tuning. The most

well-known o these changes is the removal o PermGen andthe introducion o Metaspace.

Fork/Join Speed Improvements: The ork/join ramework

was new in Java 7, and was the latest eort to simpliy

concurrent programming using the JVM. A lot o work

went into improving it urther or Java 8. Fork/join is now

the ramework thats used under the covers or parallel

operaions in the Streams API(more on this later).

In addiion, there are plenty more changes in Java 8 to

support concurrency, and Oracle has summarized some o

the perormance improvements in JDK 8.

FEWER LINES OF CODE

Java is regularly accused o being heavy on boilerplate code.

Java 8 addresses some o these issues by embracing a more

uncional style or the new APIs, ocusing on what you want

to achieve and not howto do it.

LAMBDA EXPRESSIONS

Lambda expressionsin Java 8 are not just syntacic sugar

over Javas exising anonymous inner classesthe pre-

QUICK VIEW

01In many cases, Java 8 will improve

application performance without any

specific work or tuning.

02Lambda expressions, the StreamsAPI, and new methods on existing

classes are some of the key productivity

improvements.

03Java 8s new Optionaltype gives

developers significant flexibility when

dealing with null values, reducing the

likelihood ofNullPointerExceptions
http://www.dzone.com/guideshttps://www.java.com/en/download/faq/java_7.xmlhttps://dzone.com/articles/hashmap-performancehttps://dzone.com/articles/java-8-permgen-metaspacehttps://dzone.com/articles/java-8-permgen-metaspacehttps://docs.oracle.com/javase/tutorial/essential/concurrency/forkjoin.htmlhttps://dzone.com/articles/java-7-vs-java-8-performancehttp://www.oracle.com/technetwork/articles/java/ma14-java-se-8-streams-2177646.htmlhttp://docs.oracle.com/javase/8/docs/technotes/guides/concurrency/changes8.htmlhttp://docs.oracle.com/javase/8/docs/technotes/guides/concurrency/changes8.htmlhttps://blogs.oracle.com/thejavatutorials/entry/learn_more_about_performance_andhttps://docs.oracle.com/javase/tutorial/java/javaOO/lambdaexpressions.htmlhttps://docs.oracle.com/javase/tutorial/java/javaOO/lambdaexpressions.htmlhttps://blogs.oracle.com/thejavatutorials/entry/learn_more_about_performance_andhttp://docs.oracle.com/javase/8/docs/technotes/guides/concurrency/changes8.htmlhttp://docs.oracle.com/javase/8/docs/technotes/guides/concurrency/changes8.htmlhttp://www.oracle.com/technetwork/articles/java/ma14-java-se-8-streams-2177646.htmlhttps://dzone.com/articles/java-7-vs-java-8-performancehttps://docs.oracle.com/javase/tutorial/essential/concurrency/forkjoin.htmlhttps://dzone.com/articles/java-8-permgen-metaspacehttps://dzone.com/articles/java-8-permgen-metaspacehttps://dzone.com/articles/hashmap-performancehttps://www.java.com/en/download/faq/java_7.xmlhttp://www.dzone.com/guides




Java 8 method o passing behavior around. Lambdas take

advantage o Java 7s under-the-hood changes, so they

perorm well. To see examples o where using lambda

expressions can simpliy your code, read on.

NEW METHODS ON OUR FAVORITE COLLECTIONS

While Lambdas and Streams (which well cover next) are

probably the top two selling points o Java 8, whats less well-

known is that changes in Java 8have allowed the language

developers to add new methods to exising classes without

compromising backwards compaibility. The result is that

the new methods, combined with lambda expressions, allow

us to drasically simpliy our code. Take, or example, the

common case o figuring out i an element already exists in

a Mapand creaing a new one i not. Beore Java 8, you might

write something like:

private finalMap

customers= newHashMap();

publicvoidincrementCustomerOrders(CustomerId

customerId) {

Customer customer =customers.

get(customerId);

if(customer==null) {

customer= newCustomer(customerId);

customers.put(customerId, customer);

}

customer.incrementOrders();

}

This operaion o check i the item is in the map; i not,

create it and add it is so common that theres a new method

on Mapto support it: computeIfAbsent. This method takes as

its second argument a lambda that states how to create the

missing item:


customerId) {

Customercustomer =customers.

computeIfAbsent(customerId,

id -> newCustomer(id));


}

In act, theres another new eature in Java 8 called method

reerencesthat makes this even shorter:


customerId) {

Customercustomer =customers.

computeIfAbsent(customerId, Customer::new);


}

Mapand Listboth have new methods in Java 8. Its worth

checking them out to see how many lines o code they can

save you.

STREAMS API

The Streams API gives you lexibility to query and

manipulate your data. This is a powerul tool. Check outsome o the ariclesor books on the subject or a more

complete view. Building luent queries or your data is

interesing in a Big Data world, but is just as useul or

common operaions. Lets say, or example, that you have a

list o books and you want to get a list o unique authors or

these books, in alphabeical order:

publicList

getAllAuthorsAlphabetically(Listbooks)

{

Listauthors = new ArrayList();

for(Bookbook :books) {

Authorauthor =book.getAuthor();

if(!authors.contains(author)) {

authors.add(author);

}

}

Collections.sort(authors, new

Comparator() {

publicintcompare(Authoro1, Author

o2) {

returno1.getSurname().

compareTo(o2.getSurname()); }

});

returnauthors;

}

In the code above, we first iterate through the list o books,

adding the books author to the author list i it hasnt seen it

beore; then we sort the authors alphabeically by surname.

This is exactly the sort o operaion that streams have been

designed to solve elegantly:

Generally speaking,

applications that have moved

to Java 8 see some sort of

speed improvement without

any specific work or tuning.
http://www.dzone.com/guideshttps://docs.oracle.com/javase/tutorial/java/javaOO/lambdaexpressions.htmlhttps://docs.oracle.com/javase/tutorial/java/IandI/defaultmethods.htmlhttps://docs.oracle.com/javase/tutorial/java/javaOO/methodreferences.htmlhttps://docs.oracle.com/javase/tutorial/java/javaOO/methodreferences.htmlhttp://www.oracle.com/technetwork/articles/java/ma14-java-se-8-streams-2177646.htmlhttp://www.oracle.com/technetwork/articles/java/ma14-java-se-8-streams-2177646.htmlhttps://docs.oracle.com/javase/tutorial/java/javaOO/methodreferences.htmlhttps://docs.oracle.com/javase/tutorial/java/javaOO/methodreferences.htmlhttps://docs.oracle.com/javase/tutorial/java/IandI/defaultmethods.htmlhttps://docs.oracle.com/javase/tutorial/java/javaOO/lambdaexpressions.htmlhttp://www.dzone.com/guides


8/378



publicList


{

returnbooks.stream()

.map(book ->book.

getAuthor())

.distinct()

.sorted((o1, o2) ->

o1.getSurname().compareTo(o2.getSurname()))

.collect(Collectors.toList());

}

Not only is this ewer lines o code, its arguably more

descripivea developer coming to this code later can read

it and understand that 1) its geting authors rom the books,

2) its only interested in unique authors, and 3) the list that is

returned is sorted by author surname. Combine the Streams

API with other new eaturesmethod reerencesand

new methods on Comparatorand you get an even more

succinct version:

publicList


{

returnbooks.stream()

.map(Book::getAuthor)

.distinct()

.sorted(Comparator.

comparing(Author::getSurname))

.collect(Collectors.toList());

}

Here its even more obvious that the sorted method orders

by the authors surname.

EASY TO PARALLELIZEWe spoke about better out-o-the-box perormance, and

in addiion to those earlier menioned eatures, Java 8

can explicitly make use o more CPU cores. By simply

replacing the method streamin the examples above with

parallelStream , the JVM will split the operaion into

separate jobs and use ork/jointo run them on muliple

cores. However, parallelizaion is not a magic incantaion

to make everything aster. Doing operaions in parallel

always requires more workspliting up operaions and

recombining resultsand will thereore not always take

less ime. But this opion is very interesing or areas that

are suitable or parallelizaion.

MINIMIZE NULL POINTERSAnother new eature o Java 8 is the new Optionaltype.

This type is a way o explicitly staing I might have a value,

or I might be null. Which means an API can now be explicit

about either returning values that might be null vs. values

that will a lways be non-null, minimizing the chances orunning into a NullPointerException.

Whats nice about Optionalis the way you tell it to deal

with nulls. For example, i were looking or a paricular

book in a list, t he new findFirst()method returns an

Optional, which tells us its not guaranteed to find a value.

Given this opional value, we can then decide what to do i

its null. I we wanted to throw a custom Excepion, we can

use orElseThrow :

publicBookfindBookByTitle(Listbooks,

Stringtitle) {

OptionalfoundBook =books.stream()

.filter(book ->book.getTitle().

equals(title))

.findFirst();

returnfoundBook.orElseThrow(() -> new

BookNotFoundException(Didnot find book with

title +title));

}

or you could return some other book:

returnfoundBook.orElseGet(() ->

getRecommendedAlternativeBook(title));

Or we could return an Optionalso that callers o the

method can make their own decision on what to do i the

book is not ound.

IN SUMMARYJava 8 was a big release or Java, with syntax changes, new

methods and types, and under-the-cover changes that

will help your applicaion even i you dont use the new

language eatures. Java 7 is no longer supported by Oracle,

so organizaions are being pushed to migrate to Java 8. The

good news is that Java 8 has many benefits or your business

your exising applicaion, and or developers looking to

improve their producivity.

New methods, combined with

lambda expressions, allow us to

drastically simplify our code.

TRISHA GEEhas developed Java applications for a range ofindustries, including nance, manufacturing, technology; open source

and non-protfor companies of all sizes. She has expertise in Java

high-performance systems, and is passionate about enabling developer

productivity. Trisha is a Developer Advocate for JetBrains, a leader of the

Sevilla Java User Group, a key member of the London Java Community,

a MongoDB Master, and a Java Champion.
http://www.dzone.com/guideshttps://docs.oracle.com/javase/tutorial/java/javaOO/methodreferences.htmlhttps://docs.oracle.com/javase/8/docs/api/java/util/Comparator.html#comparing-java.util.function.Function-https://docs.oracle.com/javase/tutorial/collections/streams/parallelism.htmlhttps://docs.oracle.com/javase/tutorial/collections/streams/parallelism.htmlhttps://www.java.com/en/download/faq/java_7.xmlhttps://www.java.com/en/download/faq/java_7.xmlhttps://docs.oracle.com/javase/tutorial/collections/streams/parallelism.htmlhttps://docs.oracle.com/javase/tutorial/collections/streams/parallelism.htmlhttps://docs.oracle.com/javase/8/docs/api/java/util/Comparator.html#comparing-java.util.function.Function-https://docs.oracle.com/javase/tutorial/java/javaOO/methodreferences.htmlhttp://www.dzone.com/guides



DZONES 2015 GUIDE TO THE JAVA ECOSY

DevOpsdzone.com/devops

DevOps is a cultural movement, supported by

exciting new tools, that is aimed at encouraging

close cooperation within cross-disciplinary teams

of developers and IT operations/system admins.

The DevOps Zone is your hot spot for news and

resources about Continuous Delivery, Puppet, Chef,

Jenkins, and much more.

diving deeperTOP 10 #JAVA TWITTER FEEDS

DZONE JAVA ZONESWeb Dev

dzone.com/webdev

Web professionals make up one of the largest sections

of IT audiences; we are collecting content that helps web

professionals navigate in a world of quickly changing languageprotocols, trending frameworks, and new standards for user

experience. The Web Dev Zone is devoted to all things web

developmentand that includes everything from front-end user

experience to back-end optimization, JavaScript frameworks,

and web design. Popular web technology news and releases will

be covered alongside mainstay web languages.

TOP JAVA

WEBSITES

TOP JAVA

TUTORIALS

TOP JAVA

REFCARDZ

ProgramCreekprogramcreek.com

IBM DeveloperWorksibm.com/developerworks/java

Baeldungbaeldung.com

Vogellavogella.com

Programming By Doingprogrammingbydoing.com

Java (Beginner)Programming Tutorialsbit.ly/YouTubeJava

Core JavaGives you an overview of key aspects of the Java language and

references on the core library as well as the most commonly

used tools.

Java Performance OptimizationCovers JVM internals, class loading, garbage collection,

troubleshooting, monitoring, concurrency, and more.

Getting Started with ScalaCovers creating a new Scala project, a tour of Scalas features, an

introduction to classes and objects within Scala, and much more.

Javadzone.com/java

The largest, most active Java developer

community on the web. With news and

tutorials on Java tools, performance tricks,

and new standards and strategies that

keep your skills razor-sharp.

@MREINHOLD @SPRINGROD

@TRISHA_GEE @ADAMBIEN

@LUKASEDER @PETERLAWREY @REZA_RAHMAN

@JODASTEPHEN @JBONER @JSTRACHAN

INTO THE JAVA ECOSYSTEM
http://bit.ly/1iQL31ohttp://bit.ly/1OyRMK2http://www.programcreek.com/http://www.ibm.com/developerworks/javahttp://www.baeldung.com/http://www.vogella.com/http://www.programmingbydoing.com/http://bit.ly/YouTubeJavahttp://bit.ly/1iAFkfkhttp://bit.ly/1LSztJHhttp://bit.ly/1PCNz5Ahttp://bit.ly/1iQKSTChttp://www.twitter.com/mreinholdhttp://www.twitter.com/springrodhttp://www.twitter.com/trisha_geehttp://www.twitter.com/AdamBienhttp://www.twitter.com/lukasederhttp://www.twitter.com/PeterLawreyhttp://www.twitter.com/reza_rahmanhttp://www.twitter.com/jodastephenhttp://www.twitter.com/jbonerhttp://www.twitter.com/jstrachanhttp://bit.ly/1LSztJHhttp://www.twitter.com/jodastephenhttp://www.twitter.com/trisha_geehttp://www.twitter.com/reza_rahmanhttp://www.twitter.com/jstrachanhttp://www.twitter.com/jbonerhttp://www.twitter.com/AdamBienhttp://www.twitter.com/PeterLawreyhttp://www.twitter.com/springrodhttp://www.twitter.com/lukasederhttp://www.twitter.com/mreinholdhttp://bit.ly/YouTubeJavahttp://www.programmingbydoing.com/http://www.vogella.com/http://www.baeldung.com/http://www.ibm.com/developerworks/javahttp://www.programcreek.com/http://bit.ly/1iAFkfkhttp://bit.ly/1PCNz5Ahttp://bit.ly/1iQL31ohttp://bit.ly/1OyRMK2http://bit.ly/1iQKSTC


10/3710


http://www.dzone.com/guideshttp://bit.ly/1FAKsLmhttp://www.dzone.com/guides




Generaions o developers have created muliple tools to

automaically ensure code quality, yet none o these tools

can guarantee you bug-ree code. Whatever programming

language you use, however small or large your project,

problems o various severity pop up here and there that can

only be spotted by a human. It is becoming clearer to many

o us that building peer code review into the development

process is crucial to having better quality code.

To get the most value out o code review, there are our key

aspects one needs to have in mind:

Code review does not mean that other ways o ensuring

code quality should be neglected. You sill need to write

tests, use staic code analysis, ormatter, spellchecker, etc.

Dont waste your reviewers ime on bugs that your IDE

can find or can be otherwise automated.

Distracion is the archenemy o code review; you need

to be able to ocus. Pick a code review tool that doesntspam your mailbox with noificaions and has a clean,

uncluttered UI.

Code changes are best understood in context. Try toprovide meaningul comments in your own code oryour reviewers, and use a tool that lets reviewers inspect

changes as i they were working in their IDEs.

Look or paricular things as you review, be it businesslogic, security issues, SOLID principles, or somethingelse. Scanning or general problems is surely helpul, but

when given a chance to apply your experise, take it.

The right code review tool is hal the job. The rest is up

to you!

WRITTEN BY MARIA KHALUSOVAPRODUCT MARKETING MANAGER, JETBRAINS

Getting theMost Out of

Code Review

Dont waste your reviewers time on

bugs that your IDE can find or can be

otherwise automated.

Upsource is a smart, lightweight code review and collaboraion tool that provides

unique IDE-level Java code insight.

BLOG blog.jetbrains.com/upsource WEBSITE jetbrains.com/upsourceTWITTER @upsource_jb

Upsource by JetBrains

CASE STUDY

Our team at Zando.co.za implemented a code review culture to

improve the way we developed and maintained our eCommerce

platorm, and we were looking or the right tool: ater watchingthe demo videos or Upsource on the day o its release, we

installed it r ight away. As a team, weve come to know and trust

the JetBrains brand and were excited to experiment with this

new shiny toy! Installaion was extremely simple, and we were

up and runn ing very quickly. JetBrains put a lot o thought into

2.0, and Upsource now boasts a lot o the eatures weve been

waiing or since we began code reviewing intensely. Many

thanks to JetBrains and the Upsource team or a great piece o

sotware t hat gets things done, instead o geting in our way!

STRENGTHS

Major VCSs are supported: Git, Mercurial , Subversion,

and Perorce

Upsource eatures Java code insight that helps you conduc

code review aster and more easily

With the IDE integraion plugin, you do not have to leaveyour IDE to perorm a review

Comprehensive repository browsing and powerul search

NOTABLE CUSTOMERS

CATEGORY

Code Review

NEW RELEASES

Semi-Annual

OPEN SOURCE?

No

Companies all over the world, large and small, trust team

collaboraion tools rom JetBrains.

SPONSORED OPINION
http://bit.ly/1GT3GXAhttp://bit.ly/1GT3GXAhttp://bit.ly/1GpBjFrhttp://bit.ly/1GpBjFrhttp://bit.ly/1GpBjFrhttps://twitter.com/upsource_jbhttps://twitter.com/upsource_jbhttps://twitter.com/upsource_jbhttp://bit.ly/1GpBjFrhttps://twitter.com/upsource_jbhttp://bit.ly/1GT3GXA


12/3712



Architectures based on microservices introduce new

challenges for architects and developers. An ever-

increasing list of languages and tools brings with it

the capabiliies to conquer this challenge. Java is no

excepion. This aricle explores diferent approaches to

building microservices using the Java ecosystem.

INTRODUCTIONThis aricle does notdiscuss whether microservices are good orevil, norwhether you should design your app or microservicesupront or extract the ser vices as they emerge rom yourmonolith applicaion.

The approaches described here are not the only ones available,but they should give you a pretty good overview o severalpossibiliies. Even though the Java ecosystem is the main ocusin this aricle, the concepts should be transerrable to otherlanguages and technologies.

I have named the approaches in this aricle container-less, self-contained, and in-container. These terms may not be enirelyestablished, but they ulfill their purpose here to diereniatethe approaches. I will describe what each means in t he secionsthat ollow.

CONTAINER-LESSIn the container-less approach, the developer treats everythingon top o the JVM as a part o the applicaion.

The container-less approach enables so-calledsingle JARdeployment(also called a fat JAR deployment). This means that

the applicaion, with all its dependencies, is packaged as a singleJAR file and can be run as a standalone Java process.

$ java -jar myservice.jar

One advantage o this approach is that it is extremely easy tostart and stop services as needed when scaling up or down.Another advantage is convenient distribuion. You just need topass one JAR file around.

A downside o this approach is library compaibility. You are

on your own or things l ike transacion support, or you needto bring in a third party l ibrary that provides support or thisscenario. Later oni you need support or something else,say persistenceyou may need to fight compaibility issuesbetween the libraries.

SELF-CONTAINEDAnother variant o single JAR deployment is building yourservices with an embedded ramework. In this approach, theramework provides implementaions o the services neededand the developer can choose which to include in the service.

You may argue that this is exactly the same as the container-

QUICK VIEW

01Strategies for building microservices in

the Java ecosystem include container-less,

self-contained, andin-container.

02Container-less microservices package the

application, with all its dependencies, into a

single fat JAR file.

03Self-contained microservices also

package a single fat JAR, but these also

include an embedded framework with

optional third-party libraries that will be

compatible.

04In-container microservices package an

entire Java EE container and its service

implementation in a Docker image.

First Steps

in JavaMicroservicesBY IVAR GRIMSTAD

SECURITY PERSISTENCE MESSAGING TRANSACTIONS INTEGRATION PRESENTATION

APPLICATION

JV M

OPERATING SYSTEM

HARDWARE




lesssoluion, but I like to dis inguish them here since theself-containedapproach actually gives you a set o third partylibraries that you knoware compaible.

This approach can involve tools like Spring Bootand WildlySwarm.

SPRING BOOT

Spring Bootand the Spring Cloudprojects have excellentsupport or building microservices in Java. Spring Boot allowsyou to pick and choose various par ts o the Spring ecosystem,as well as popular external tools and then package them alongwith your applicaion in a JAR file. Spring Iniializrallows

you to do this with a simple checkbox list orm. A simple HelloWorldservice is shown in this example:

GIST SNIPPET: bit.ly/1YgUO9c

WILDFLY SWARM

A Java EE counterpart to Spring Boot is WildFly Swarm. Itenables you to pick and choose which parts o the Java EEspecificaion you need and package them and your applicaionin a JAR file. The Hello World example looks like this:

GIST SNIPPET: bit.ly/1LADHW9

The advantage o theself-containedapproach is that you get toselect only what you need in order or the service to run.

One disadvantage o this approach is that the configuraion isa little more complex and the resuling deliverable JAR file is abit bigger since it builds in the required container capabiliiesin the actual service.

IN-CONTAINERWhile it seems li ke a lot o overhead to require an enire JavaEE container to be able to deploy a microservice, keep in mindthat some developers argue that the the micro in microservicedoes not necessarily mean that the service is small or simple.

In these cases it may seem appropriate to treat the Java EEcontainer as the required platorm. Thus, the only dependencyyou need is the Java EE API. Note that the dependency isprovided since the implementaion is providedby the container.That means that the resuling WAR file is extremely lean.

The implementaion o the service is the same as the Wild lySwarm example above. See it here:

GIST SNIPPET: bit.ly/1ik7Hya

The advantage o this approach is that the container providestested and verified implementaions o standard uncionalitythrough standard APIs. Thus, you as a developer can ocusenirely on the business uncionality and leave the plumbingout o the applicaion source. Another advantage o thisapproach is that the actual applicaion code does not depend

on the Java EE applicaion server it is deployed to, whether it isGlassFish, WildFly, WebLogic, WebSphere, or any other Java EEcompaible implementaions.

The disadvantage is that you need to deploy the service into acontainer and thus increase the complexity o the deployment.

DOCKER

This is where Dockercomes in. By packaging the Java EEContainer and the service implementaion in a Docker image,you achieve more or less the same result as you would with asingle JAR deployment. The di erence is that now the serviceis contained in a Docker image and not a JAR file.

Dockerfile

FROM jboss/wildfly:9.0.1.Final

ADD myservice.war /opt/jboss/wildfly/standalone/

deployments

The service is started by staring the Docker image in theDocker engine.

$ docker run -it -p 8081:8080 myorganization/myservice

SNOOP

The observant reader may have noiced the

@EnableEurekaClientannotaion in the Spring Boot codesnippet rom beore. This annotaion registers the service with

Eureka,making it discoverable by service consumers. Eurekais a part o the Spring Cloud bundle and is an extremely easy-to-use and configure service discovery soluion.

Java EE does not oer this uncionality out o the box,but there are several open-source soluions available. Onesuch soluion is Snoop, which uncions in a simi lar way toEureka. The only thing needed to make a Java EE microserviceavailable or service lookup is the @EnableSnoopClientannotaion as shown in this example:

GIST SNIPPET: bit.ly/1Kp2PP1

CONCLUSIONJava is an excellent choice when building microservices. Any

o the approaches described in this ar icle will get things done.The most appropriate method or your paricular case dependson the requirements o the service. For simpler services, acontainer-less orself-containedservice is the better choice, butmore advanced services may be aster and easier to implementwith the power o an in-containerimplementaion. Either way,Java isa proven ecosystem or implemening microservices.

IVAR GRIMSTADis an experienced software architect andconference speaker focusing on Enterprise Java. He is a member of the

JCP and currently works on JSR 371 (MVC 1.0), JSR 375 (Java EE Security

API), and JSR 368 (JMS 2.1). Working as a Java developer since the

languages very beginning, he has built applications using everything from

JavaEE to Spring and a variety of other open-source products.

APPLICATION

JV M

SE CU RITY PERSISTENCE MESSAGING TRANSACTIONS INTE GRATION PRESENTATION

OPERATING SYSTEM

HARDWARE

APPLICATION

JAVA EE CONTAINER

JV M

OPERATING SYSTEM

HARDWARE
http://www.dzone.com/guideshttp://projects.spring.io/spring-boot/http://projects.spring.io/spring-cloud/https://start.spring.io/http://bit.ly/1YgUO9chttp://wildfly.org/swarm/http://bit.ly/1LADHW9http://bit.ly/1ik7Hyahttps://glassfish.java.net/http://wildfly.org/http://www.oracle.com/us/products/middleware/cloud-app-foundation/weblogic/overview/index.htmlhttp://www.ibm.com/software/webspherehttps://www.docker.com/https://github.com/Netflix/eureka/wiki/Eureka-at-a-glancehttps://github.com/Netflix/eureka/wiki/Eureka-at-a-glancehttps://github.com/ivargrimstad/snoophttp://bit.ly/1Kp2PP1http://bit.ly/1Kp2PP1https://github.com/ivargrimstad/snoophttps://github.com/Netflix/eureka/wiki/Eureka-at-a-glancehttps://www.docker.com/http://www.ibm.com/software/webspherehttp://www.oracle.com/us/products/middleware/cloud-app-foundation/weblogic/overview/index.htmlhttp://wildfly.org/https://glassfish.java.net/http://bit.ly/1ik7Hyahttp://bit.ly/1ik7Hyahttp://bit.ly/1LADHW9http://wildfly.org/swarm/http://bit.ly/1YgUO9chttp://bit.ly/1YgUO9chttps://start.spring.io/http://projects.spring.io/spring-cloud/http://projects.spring.io/spring-boot/http://www.dzone.com/guides


14/37

Source:TechvalidatesurveyofRiverbedusersofSteelC

entralandPerformanceManagement

2015 Riverbed Technology. All rights reserved. Riverbed and any Riverbed product or service name or logo used herein are trademarks of Riverbed Technology. All other trademarks usedherein belong to their respective owners. The trademarks and logos displayed herein may not be used without the prior written consent of Riverbed Technology or their respective owners.

How to detect & fix

problems 10x faster

Were now able to look inside of the

developers code without having to

modify the code while its running in our

production environment. Thats fantastic.

I cant imagine someone running a site ofany real size without this capability.

AppInternals has drastically

reduced the amount of timerequired to pinpoint issues and

allowed us to pinpoint wherethe problems exist.

Eric McCraw, Global Web Systems Manager for IT,National Instruments

Eric Saxe, IT Manager, Asurion

Top 5 benefits cited by users

Challenges

Improved appperformance

Reduced timespent in war

rooms

Reducedcost of

fixing appsImprovedapp user

satisfaction

Fasterdevelopment ofnew apps and

features

TechValidate ID: 958-DAB-5CA

1 2

3

4

5

IDG, 2013 Gartner, 2012 Forrester, 2013

of ITorganizations

suffer fromslow apps

75%

learn aboutproblemsfrom end

users

70%

performanceissues take morethan a month to

resolve

31% Performance

Transactions

User Experience

Application Map

1:1 fit

Take control of app performance with

SteelCentral AppInternals

Try AppInternals today atwww.appinternals.com
http://bit.ly/1Ng3QNyhttp://bit.ly/1Ng3QNy




Applicaion perormance equals business perormance. Apps

have transormed how businesses operate, have increased

producivity, and have enabled numerous innovaive

markets. Developers make this possible.

As indispensable as applicaions have become, users expect

more eatures as needs evolve and depend on developers to

deliver these requently, without compromising reliability or

around-the-clock availability. Failing to do so will drive users

to compeitors or alternaives.

How do developers expedite release cycles and maximize

ime spent on new eatures as opposed to maintaining

exising code? It comes down to two pracices:

ENSURE RELEASE QUALITY PRIOR TO RELEASE

Ideniying all erroneous code during tesing helpsdevelopers fix issues early. Detailed transacion tracing canhelp QA expose and diagnose bottlenecks so that developershave the inormaion they need to resolve them quickly. Thishelps make releases producion ready.

RESOLVE ERRORS IN PRODUCTION ASAP

Monitoring apps in producion ensures that you are the firstknow when problems strike. Recording every transacionrom user to back-endalong with its associated systemmetrics and call tree detailshelps developers reconstructincidents and eliminate the root cause quickly.

Helping developers stay ocused on new eatures is importanor businesses to stay compeiive. Applicaion perormancemonitoring products arm them with the diagnosics theyneed to minimize the ime spent fixing bugs, maximizerelease quality, and keep them ocused on deliveringbusiness-impacing applicaions.

WRITTEN BY KRISHNAN BADRINARAYANANSR. PRODUCT MARKETING MANAGER, RIVERBED TECHNOLOGY

Developing andDelivering High-

PerformingApplicationsWE R E NOW A BLE TO LOOK IN SIDE OF T HE DEV ELOPERS

CODE WITHOUT HAVING TO MODIFY THE CODE WHILE

ITS RUNNING IN OUR PRODUCTION ENVIRONMENT. THATS

FANTASTIC. I CANT IM AGINE SOMEONE RUNNING A SITE OF

AN Y R EA L SIZ E W IT HOU T THIS C APA BILI TY.

GLOBAL WEB SYSTEMS MANAGER, NATIONAL INSTRUMENTS

Trace all transacions rom user device to back-end. Get comprehensive visibility and

analyics across apps, networks and inrastructure.

BLOG riverbed.com/blogs/ WEBSITE appinternals.comTWITTER @riverbed

SteelCentral by Riverbed Technology

CASE STUDY

The web systems team at Naional Instruments is tasked with

ensuring that its public website, www.ni.com, runs opima lly.

They used to spend 1000s o hours each year troubleshooingissues caused by newly released apps. Their inability to quickly

find causes o app perormance problems created tension between

them and the developers, and oten hurt user experience. The

web systems team now uses AppInternals to quickly diagnose

root causes o app perormance problems. Troubleshooing ime

is down by 90%, along with MTTR. Developers use AppInternals

to test their code resuling in 20% to 30% ewer issues introduced

into producion. Scheduled updates are up rom 16 to 120 per year.

ni.com is more stable and delivers better user experience.

STRENGTHS

Major VCSs are supported: Git, Mercurial , Subversion, andPerorce

Upsource eatures Java code insight that helps conductcode review aster and easier

With IDE integraion plugin, you do not have to leave your

IDE to perorm a review

Comprehensive repository browsing and powerul search

NOTABLE CUSTOMERS

CATEGORY

APM

NEW RELEASES

Quarterly

OPEN SOURCE?

No

Asurion

Naional Instruments

Shell

Visa

SPONSORED OPINION
http://bit.ly/1NMfl26http://bit.ly/1NMfl26http://bit.ly/1NMfl26http://bit.ly/1jsCKclhttp://bit.ly/1jsCKclhttp://bit.ly/1jsCKclhttp://bit.ly/1GKiAPRhttp://bit.ly/1GKiAPRhttp://bit.ly/1GKiAPRhttp://bit.ly/1jsCKclhttp://bit.ly/1GKiAPRhttp://bit.ly/1NMfl26


16/3716



Today more than ever, speed plays a larger role

in the sotware development lifecycle. We see R&D

teams who want to push code faster to producion

environments with rising complexity, and this

amplifies a vulnerability that must be addressed.

Those ew hours ater a new deployment set the tone or its

success. Every once in a while, things go wrong, no matter how

strict your tests are. When your code is out in producion and

it meets the real-world architecture and scale o your system,

with real data lowing through the applicaion, things can go

south pretty quickly. In order to be resilient and stay on top o

things, a strategy needs to be implemented that allows you to:

Ideniy when theres an error happening

Assess the errors severity to prioriize it

Draw out the state that caused the error

Trace back and solve the root cause

Deploy a hotfix

In this ar icle well cover some o the most useul pracices toallow you to assemble a ime-criical response and weaponizeyour applicaion.

THE CASE FOR DISTRIBUTED LOGGINGWith producion environments spread across muliple nodes

and clusters, its easy or a transacion that starts on one

machine or service to cause an error someplace else. When an

excepion happens, theres a need to be able to trace back this

type o distributed transacion, and the logs are oten the first

place to look or clues.

This is why, or every log line printed out, we need to be able to

extract the ull context to understand exactly what happened

there. Some data might come rom the logger itsel and the

locaion the log is created in; other data needs to be extracted

at the moment o the event. A good way to trace such errors

to their origin would be generaing UUIDs at every threads

applicaion entry point.

A useul yet underuilized eature here is using thread names

to provide a window or this precious context, right beore the

stack collapses and the data is lost. You can ormat your thread

name to something like:

Thread.currentThread().

setName(prettyFormat(threadName, getUUID(), message.

getMsgType(), message.getMsgID(), getCurrentTime()));

So instead o an anonymous name like pool-1-thread-17your

applicaion now produces smart stack traces that start this

way: threadName: pool-1-thread-17, UUID: AB5CAD, MsgType:

AnalyzeGraph, MsgID: 5678956, 30/08/2015 17:37

This works well when handling caught excepions, but what

i theres an uncaught excepion involved? A good pracice is

to set a deault uncaught excepion handler, both to cover or

that and to help extract any useul data you need. Other than

thread names, addiional places we can use to store hints about

what happened are the TLS (Thread Local Storage) and the MDC

(Mapped Diagnosic Context, which is provided by your logging

ramework). All other data gets lost as the stack rames collapse.

LEANING ON THE JVM TOOL BELTSome more complex bugs like deadlocks or heavy perormance

QUICK VIEW

01Developers need as much context information

as possible in their logs when debugging in

production. The data extraction needs to be

planned and built in before deployment to

production so that context data is not lost as the

stack frames collapse.

02For bugs like deadlocks or heavy performance

bottlenecks, jstack is a great debugging tool,

but developers need to modify it for preemptive

execution if they want to collect information at

the exact time the bug occurs.

03Java agents are an important debugging tool.

They retrieve data straight from the source in

a lightweight way, allowing access to the exact

variable values that caused each error.

04Successful production testing strategies can

include duplicating live traffic into new appversions or creating a canary server with varying

degrees of features and controls.

Production

Debugging IsNot a CrimeBY ALEX ZHITNITSKY




bottlenecks require a di erent approach. Take jstack, or

instance: a powerul tool that ships together with the JDK.

Most o you are probably already amiliar with it in some way.

Basically, jstack allows you to hook into a running process and

output all the threads that are currently running in it. It will

print each threads stack trace; rameseither Java or naive;

locks theyre holding; and all sorts o other metadata. It can

also analyze heap dumps or core dumps o processes that have

already ended. Its a longstanding and super useul toolkit.

The problem here is that jstack is mostly used in retrospect.

The condiion youre looking to debug has already happened,

and now youre let searching through the debris. The server

isnt responding, the throughput is dropping, database queries

are taking orever: a typical output would be a ew threads

stuck on some nasty database query, with no clue o how we

got there. A nice hack that would allow you to get the jstack

output where it matters most is to acivate it automaically

when things start tumbling down. For example, you can set

a certain throughput threshold and get jstack to run at the

moment it drops[1].

Combined with using smart thread names, we can now know

exactly which messages caused us to get stuck, and we can

retrace our steps back, reproduce the error, isolate it, and solve it.

USING JAVA AGENTS TO SKIP LOGGING ALTOGETHERThe next step in this process is gaining visibility into your

applicaion during runime. Logs are inspected in retrospect

and only include the inormaion that youve decided to put

there in advance. Weve seen how we can enrich them with

stateul data, but we also need a way to access the exact

variable values that caused each error to get down to the real

root cause. Java agents give us the ability to get to the data

we need straight rom the source without wriing to disk

and using huge log files, so we can extract only the data well

actually be using.

One interesing approach is using BTrace, an open-source Java

agent that hooks up to a JVM and opens up a scriping language

that lets you query it during runime. For instance, you can get

access to things like ClassLoaders and their subclasses, and load

up jstack whenever some troubled new class is instaniated. Its

a useul tool or invesigaing specific issues and requires you to

write scripts or each case you want to cover.

You could also write your own custom Java agent, just like

BTrace. One way this helped our team at Takipiwas when

a certain class was instaniaing millions o new objects

or some reason. We wrote an agentthat hooks up to the

constructor o that object. Anyime the object was allocated

an instance, the agent would extract its stack trace. Later

we analyzed the results and understood where the load was

coming rom. These kinds o problems really pique our teams

interest. On our day-to-day were building a producion grade

agentthat knows how to extract t he variable values that causeeach excepion or logged error, all across the stack trace, and

across dierent machines.

TESTING IN PRODUCTION: NOT FOR THE FAINT-HEARTEDJokes aside, tesing in producion is a serious pracice that

many companies are taking part in . They dont cancel the

tesing phase completely, but they understand that staging

environments are not enough to mimic ull-blown distributed

environments, no matter how much ime and e ort you

put into seting them up. The only real tesing takes place in

producion, with real data lowing through the system andunexpected use cases being thrown at it.

There are several approaches you can adopt or perorming

controlled producion tesing, depending on what kind o

uncionality it is that youre trying to test. One opion is

duplicaing and rouing live trafic both through the current

system and through the new version o the component that

youre tesing. This way you can see how the new component

behaves and compare it direct ly to the current version without

risking the delivery o wrong results back to the user i, or

example, its some data crunching task.

Another opion is segmening your tra fic and releasing

new eatures gradually. One way to do this is to use a canary

server, which is a single node in your system updated with

the new version youd like to roll out (just like a canary in

a coal mine). Someimes its also helpul to add more fine-

grained monitoring and logging to the canary server. Another

opion is to add more abstracion on top o the canary setup,

implemening and making use o gradual rollouts with eature

switches, or A/B tesing sma ll changes in your applicaion to

see how they impact perormance.

FINAL THOUGHTSDebugging Java applicaions in producion requires a creaive

and orward-thinking mindset. Unless we prepare our

applicaions and environment in advance, there will not

be much insight to recover ater geting hit by errors and

perormance issues.

[1] https: //github.com/takipi/jstack

ALEX ZHITNITSKYis an engineer working at Takipion a mission tohelp Java and Scala developers solve bugs in production and rid the

world of buggy software. Passionate about all things tech, he is also the

co-founder & lead of GDG Haifa, a local developer group. Alex holds a

B.Sc from the Technion, Israels Institute of Technology.

Unless we prepare our applications

and environment in advance, there

will not be much insight to recover

after getting hit by errors and

performance issues.
http://www.dzone.com/guideshttp://docs.oracle.com/javase/7/docs/technotes/tools/share/jstack.htmlhttps://github.com/takipi/jstack/https://github.com/takipi/jstack/https://kenai.com/projects/btracehttps://www.takipi.com/https://github.com/takipi/debugAgenthttp://land.takipi.com/why-takipi/http://land.takipi.com/why-takipi/https://github.com/takipi/jstackhttps://www.takipi.com/https://github.com/takipi/jstackhttps://www.takipi.com/https://github.com/takipi/jstackhttp://land.takipi.com/why-takipi/http://land.takipi.com/why-takipi/https://github.com/takipi/debugAgenthttps://www.takipi.com/https://kenai.com/projects/btracehttps://github.com/takipi/jstack/https://github.com/takipi/jstack/http://docs.oracle.com/javase/7/docs/technotes/tools/share/jstack.htmlhttp://www.dzone.com/guides


18/37

IEEE LanguageRankings 2015Methodology:

Measures Google searches, social media trends, GitHub and SO projects/tags plus

activity, and recent job postings on Career Builder and Dice.

RedMonk 15 Program.Language RankingsMethodology:

Measures the number of repositories on GitHub using each language and the number of

questions on Stack Overflow with each language tag.

05 06

1

2

3

4

5

6

7

8

9

10

2008 09 10 11 12 13 14 15

JAVASCRIPT

JAVA

RUBY

PHP

PYTHON

CSS

C++

C#

C

PERL

OBJECTIVE-C

HTMLSHELLVmLEMACS LISP

TIOBE ProgrammingCommunity Index

GitHub LanguageRankings

Javas Current Rank: #1

SOURCE tiobe.com/index.php/content/paperinfo/tpci/

JAVA

C

C++

C#

PYTHON

OBJECTIVE-C

PHP

JAVASCRIPT

PERL

2003 04 05 06 07 08 09 10 11 12 13 14 15

2006 07 08 09 10 11 12 13 14 15

1

2

3

4

5

6

7

8

9

0

1%

2%

3%

MATCHING

JOBP

OSTINGS

RANK


SOURCE github.com/blog/2047-language-trends-on-github


SOURCE bit.ly/indeed-rankings


SOURCE makingdataeasy.com/stackoverflow-trends

Methodology:

Measures language search trends on Google, Google Blogs, Amazon, Bing, Yahoo!,

Baidu, Wikipedia, and YouTube.

Indeed Job KeywordTrendsMethodology:

Measures Indeed.com job posting keyword search results for selected languages. C was

not included because there were too many false positives.

Methodology:

Measures the number of public and private GitHub repositories using each language

(excluding forks) as detected by Linguist.

Stack OverflowLanguage RankingsMethodology:

Measures the number of questions on Stack Overflow with each language tag.

01 02

03 04

JAVA C++ C# JAVASCRIPT PERL PHP PYTHON

RANK

RANKINGS

OVER

TIME

RANKINGS

OVER

TIME

CUR

RENT

RANKINGS

CUR

RENT

RANKINGS

RANKINGS

OVERT

IME

RANKINGS

OVERT

IME

RAN

K

1

2

3

4

5

6

7

8

9

10

2008 09 10 11 12 13 14 15

RUBY

JAVASCRIPT

JAVA

C#

PHP

PYTHON

C++

CSS

SQL

OBJECTIVE-C

C

JAVAJAVA C 99 9100 21

Java Popularity:By the Numbers

Anyone whos spent a little time in the software industry

knows how widespread the use of Java is, but how popular

is it exactly? There are multiple places where we can look to

help answer this question. Google searches, job postings,

open source repositories, Stack Overflow questions, and

social media are all good sources to give us a fairly accurate

picture of language popularity overall. Here are some chart

to show how amazingly popular Java is.
http://www.tiobe.com/index.php/content/paperinfo/tpci/index.htmlhttp://www.tiobe.com/index.php/content/paperinfo/tpci/index.htmlhttps://github.com/blog/2047-language-trends-on-githubhttps://github.com/blog/2047-language-trends-on-githubhttp://www.bit.ly/indeed-rankingshttp://www.bit.ly/indeed-rankingshttp://makingdataeasy.com/stackoverflow-trendshttp://makingdataeasy.com/stackoverflow-trendshttps://github.com/github/linguisthttps://github.com/github/linguisthttp://makingdataeasy.com/stackoverflow-trendshttp://www.bit.ly/indeed-rankingshttp://www.tiobe.com/index.php/content/paperinfo/tpci/index.htmlhttps://github.com/blog/2047-language-trends-on-githubhttps://github.com/github/linguist


19/3720



Two years ago, I collaborated with a few other

developers to define a new set of architectural

principles in the enterprise. It was clear at that

ime that everything from emergent deployment

environments to user expectaions to the size of

datasets had outgrown previous patterns for buildingsotware. Basically what was typically breaking

ied back to most sotware having synchronous call

request chains and poor isolaion, yielding single

points of failure and too much contenion.

In the Reacive Maniestowe created a new set o keyprinciples that described the responsiveness, resilience,elasicity, and message-driven characterisics that webelieved defined eecive applicaion architectures in themodern era o applicaions, running on everything rommobile devices to cloud-based clusters with t housands omuli-core processors.

Over the last two years, the response rom the communityhas ranged rom enthusiasic support (12,000+ signatureson the Reacive Maniesto) to eye-rol ling (my avorite,in a Slashdot comment: We want a machine that makesthings cold. We dont care how its built. Well call t his...The Rerigerator Maniesto.). Some elt that Reaciveencapsulated key attributes that had long been embraced intheir inter nal development philosophy, in the same way thatsome companies did Agile sotware development beore itwas Agile. Some elt Reacive too prescripive, while otherselt it was too generic.

SO WHAT IS REACTIVE?The Reacive Maniestos goal is to condense the knowledgearound designing highly-scalable and reliable applicaionsinto a set o our required architecture traits:

RESPONSIVE

The systemresponds in a imely manner i at all possible.Responsiveness is the cornerstone o usability and ui lity, but

more than that, responsiveness means that problems maybe detected quickly and dealt with eecively. Responsivesystems ocus on providing rapid and consistent responseimes, establishing reliable upper bounds so they deliver aconsistent quality o service. This consistent behavior in turnsimplifies error handling, builds end user confidence, andencourages urther interacion.

RESILIENT

The systemstays responsive in the ace o ailure. This appliesnot only to highly-available, mission criical systems anysystem that is not resilient will be unresponsive duringand ater ailure. Resilience is achieved by replicaion,containment, isolaion, and delegaion. Failures are contained

within each component, isolaing components rom eachother and thereby ensuring that parts o the system can ailand recover without compromising the system as a whole.Recovery o each component is delegated to another (external)component and high-availability is ensured by replicaionwhere necessary. The client o a component is not burdenedwith handling its ailures.

ELASTIC

The system stays responsive under varying workload. ReaciveSystems can react to changes in the input rate by increasingor decreasing the resourcesallocated to service these inputs.This implies designs that have no contenion points or central

QUICK VIEW

01Reactive Systems rely on a foundation

of asynchronous message-passing to

create loosely-coupled systems that areresponsive, resilient, and elastic.

02Analytics are being pushed into thestream (via Spark), which is emergingas the de facto approach for sub-second

query response times across billions of

rows of data.

03Fast Data, or streaming data, can

make systems more responsive.Reactive Streams can make Fast Data

manageable by intelligently controlling

the rate of data consumption.

04Systems become very resilient when they

are backed by a database with the fullhistory of an application, which is built

by event logging that is being used as

the Service of Record.

Reactive

Trends onthe JVMBY JONAS BONR
http://www.dzone.com/guideshttp://www.reactivemanifesto.org/http://developers.slashdot.org/story/14/09/21/0547231/reactive-development-turns-20http://www.reactivemanifesto.org/glossary#Systemhttp://www.reactivemanifesto.org/glossary%23Systemhttp://www.reactivemanifesto.org/glossary#Failurehttp://www.reactivemanifesto.org/glossary#Replicationhttp://www.reactivemanifesto.org/glossary#Isolationhttp://www.reactivemanifesto.org/glossary#Delegationhttp://www.reactivemanifesto.org/glossary#Componenthttp://www.reactivemanifesto.org/glossary#Resourcehttp://www.reactivemanifesto.org/glossary#Resourcehttp://www.reactivemanifesto.org/glossary#Componenthttp://www.reactivemanifesto.org/glossary#Delegationhttp://www.reactivemanifesto.org/glossary#Isolationhttp://www.reactivemanifesto.org/glossary#Replicationhttp://www.reactivemanifesto.org/glossary#Failurehttp://www.reactivemanifesto.org/glossary%23Systemhttp://www.reactivemanifesto.org/glossary#Systemhttp://developers.slashdot.org/story/14/09/21/0547231/reactive-development-turns-20http://www.reactivemanifesto.org/http://www.dzone.com/guides




bottlenecks, resuling in the ability to shard or replicatecomponents and distribute inputs among them. ReaciveSystems support predicive, as well as Reacive, scalingalgorithms by providing relevant live perormance measures.They achieve elasicityin a cost-eecive way on commodityhardware and sotware platorms.

MESSAGE-DRIVEN

As the oundaion to the three traits above, Reacive Systemsrely on asynchronousmessage-passingto establish a

boundary between components that ensures loose coupling,isolaion, locaion transparency, and provides the means toreiy and delegate ailuresas messages. Employing explicitmessage-passing enables load management, elasicity,and low control by shaping and monitoring the messagequeues in the system and applying back-pressurewhennecessary. Locaion-transparent messaging as a means ocommunicaion makes it possible or the management oailure to work with the same constructs and semanicsacross a cluster or within a single host. Non-blockingcommunicaion allows recipients to only consume resourceswhile acive, leading to more eficient system uilizaion.

NEW PATTERNS DRIVING REACTIVE INNOVATIONOver the last couple o years, I believe the three mostinteresing new trends driving Reacive innovaion are:

MICROSERVICES

In tradiional Java EE apps, services are written in a verymonolithic way. That ies back to a strong coupling betweenthe components in the service and between ser vices. Appservers (WebLogic, JBoss, Tomcat, etc.) are encouragingthis monolithic model. They assume that you are bundlingyour service JARs into an EAR file as a way o grouping yourservices, which you then deployalongside all your otherapplicaions and servicesinto the single running instance othe app server, which manages the service isolaion through

class loader tricks; a very ragile model.

Today we have a much more evolved oundaion orisolaionrom the ground upstaring with, or example,Docker containers, isolated all the way up through betterhardware and communicaion protocols. Im excited aboutthe Microservices momentum because it makes isolaionfirst class, which is a necessity or resilience. You cant builda Reacive system without isolaing ailures and having aseparate context outside the ailed component to react to theailure. You need isolaion in order to avoid cascading ailures.

FAST DATATHE WORLD IS GOING STREAMING

The whole movement towards Fast Data and real-ime data

requires closed eedback loops or geting data into and outo the system. T he benefit o Fast Data is that you get systemsthat are more responsive and adapive, allowing you to eedthe results o real-ime data processing back into the runningsystem, which allows it to react to change. This capabilitycan also be used to make these systems more resilient andscalable, but with reduced complexity.

One o the early pitalls or streaming data in Fast Datascenarios, or example, was the lack o back-pressure. I a

processing stage produced data aster than the next stage couldconsume that data, it would lead to a ailure in the consumer,which would cause cascading ailures throughout the enireprocessing pipeline. Reacive Streamsaddress that problemby bringing back-pressure to streams to control the rate o dataconsumpion. In general, the back-end systems that scale BigData and IoT are a perect fit or Reacive architectures.

EVENT LOGGING AS THE SERVICE OF RECORD

Another area in which Im seeing a lot o Reacive innovaion,specifically on the JVM, is when event logging is being usedas the Ser vice o Record. In event logging, each state changeto the applicaion is materialized as an event in the log. What

you get is a database with the ull history o the applicaion;a database o acts, rather than the tradiional SQL databaseapproach that only works with a cache o the subset o thelog, as Pat Helland aptly put it. I your durable state is basedon an event log, it can be easi ly replicated and replayedsomewhere else to bring t he system or component up to speedwherever it is. This is a great pattern or ailure handling indistributed stream processingi one thing ails it can bebrought back up to speed and coninue. Architectural patternsmaking use o the event log include Event Sourcing and CQRSThis way o thinking about durable state works very well withMicroservices, where each ser vice can have its own isolated,strongly consistent, event log-based, durable storage thatrelies on eventual consistency between t he services or scale

and availability. Fast Data is the oundaion or this durablestate and streaming architecture.

REACTIVE EVOLVING FROM PRINCIPLES TOIMPLEMENTATION PATTERNSI believe that Reacive is on a similar arc to the one the Agilemethodology ollowed. A lot o companies were using Agilemethodologies without calling it Agile years beore the termwas coined. But calling it Agile made it easier to talk aboutand communicate the philosophy, so the Agile Maniestowascreated. When this set o principles became more amiliar in thedeveloper community, you started to see the adopion o Scrumor disilling the essence o the processes and XP or supporingprogramming principles and pracices like test-drivendevelopment, pair programming, and coninuous integraion.

Today were seeing the same shi t or Reacive. The coremessage o Reacive is aimed at core principles rather thantools and techniques. But Microser vices, Fast Data, and EventLogging are great examples o how implementaion patternswithin the Reacive movement are staring to get moredefiniion and momentum.

JONAS BONRis the co-founder and CTO of Typesafe. He is also theinventor of Akka, a JVM-based toolkit and runtime that uses the actor

model to build concurrent, distributed applications. Jonas is a Java

Champion and a co-author of the Reactive Manifesto.

What was typically breaking

tied back to most software having

synchronous call request chains and

poor isolation, yielding single points

of failure and too much contention.
http://www.dzone.com/guideshttp://www.reactivemanifesto.org/glossary#Elasticityhttp://www.reactivemanifesto.org/glossary#Asynchronoushttp://www.reactivemanifesto.org/glossary#Message-Drivenhttp://www.reactivemanifesto.org/glossary#Location-Transparencyhttp://www.reactivemanifesto.org/glossary#Failurehttp://www.reactivemanifesto.org/glossary#Failurehttp://www.reactivemanifesto.org/glossary#Non-Blockinghttp://www.reactivemanifesto.org/glossary#Resourcehttp://www.reactivemanifesto.org/glossary%23Back-Pressurehttp://www.reactive-streams.org/http://blogs.msdn.com/b/pathelland/archive/2007/06/14/accountants-don-t-use-erasers.aspxhttp://www.agilemanifesto.org/http://www.agilemanifesto.org/http://blogs.msdn.com/b/pathelland/archive/2007/06/14/accountants-don-t-use-erasers.aspxhttp://www.reactive-streams.org/http://www.reactivemanifesto.org/glossary%23Back-Pressurehttp://www.reactivemanifesto.org/glossary#Resourcehttp://www.reactivemanifesto.org/glossary#Non-Blockinghttp://www.reactivemanifesto.org/glossary#Failurehttp://www.reactivemanifesto.org/glossary#Failurehttp://www.reactivemanifesto.org/glossary#Location-Transparencyhttp://www.reactivemanifesto.org/glossary#Message-Drivenhttp://www.reactivemanifesto.org/glossary#Asynchronoushttp://www.reactivemanifesto.org/glossary#Elasticityhttp://www.dzone.com/guides


21/3722



Constantly monitoring

your applications soyou dont have to.

New Relics SaaS-based Application

Performance Monitoring helps you

build, deploy, and maintain great web

and mobile software.

Start your free trial now.

newrelic.com/java

Application Response Times

Error Tracking & Analytics

Transaction Tracing

Thread Profiling

JVM Performance Analysis

Deployment Comparisons

2008-15 New Relic, Inc. All rights reserved.
http://bit.ly/1Ng3GWchttp://bit.ly/1Ng3GWchttp://bit.ly/1Ng3GWchttp://bit.ly/1Ng3GWchttp://bit.ly/1Ng3GWchttp://bit.ly/1Ng3GWchttp://bit.ly/1Ng3GWchttp://bit.ly/1Ng3GWchttp://bit.ly/1Ng3GWchttp://bit.ly/1Ng3GWchttp://www.newrelic.com/javahttp://www.newrelic.com/javahttp://bit.ly/1Ng3GWchttp://bit.ly/1Ng3GWchttp://bit.ly/1Ng3GWchttp://bit.ly/1Ng3GWchttp://bit.ly/1Ng3GWchttp://bit.ly/1Ng3GWchttp://bit.ly/1Ng3GWchttp://www.newrelic.com/javahttp://bit.ly/1Ng3GWchttp://www.dzone.com/guides




Developer aigue, that rustrated exhausion you eel

trying to keep up with the constant lood o new languages,

libraries, rameworks, platorms and programming models,

is a real problem. While theres nothing wrong with trying

to stay up-to-date, a number o downsides work against you:

loss o producivity, lost cycles spent on immature tools, short

hal-lie and the risk o making bad bets.

This problem has been around since the dawn o computers,

but its geting worse. Causes include the push or ull stack

developers who must do more with less, the rising numbero open source projects, the prolieraion o new types o

hardware and the pressure to introduce the new in order to

stay ahead o the compeiion.

How can you possibly keep up? Perhaps the best way is to take

it slowly: integrate the eatures o new technologies as you

need them. In my experience, the need to use a technology is

the best reason to start learning about it and the best wayto retain the skills once learned. Jumping in and learninga new technology or its own sake (aka warehousing) is

popular, but much less eecive.

With gradual introducion to new technologies and a ocus othe aspects that can help solve immediate problems, you canstay producive while sill using amiliar tools, and introduc

new stu one eature at a ime. While not possible in all

cases, i your amiliar platorm is Java or .NET, and the newtechnology is .NET or Java-based, JNBridgePro can help.

JNBridge has examplesthat show how a amiliar legacytechnology like .NET or Java can be integrated with a newer

technology like Hadoop, Groovy, Python or Clojure. The

examples ocus on how you can bring in new technologiesat your own pace, without having to assimilate the enirety

o a new language, API, or platorm. Our goal is to help youavoid developer aigue while keeping current on the ever-

increasing number o new technologies being introduced.

Developer fatigue is frustrating andcostly. Stay productive using famili

guide to the java ecosystem

Documents