Presented at JAX London 2015. The last few years have seen a huge growth in the usage of JavaScript, to the extent that it is often reported to be the #1 programming language in use today. Additionally, the arrival of server-side JavaScript through frameworks such as Node.js and Ringo.js, and JavaScript on the JVM through Nashorn and Avatar.js, means that enterprise web applications written in JavaScript are not just a possibility—but a reality for companies such as LinkedIn, eBay, Yahoo, ADP and Dow Jones. This session will compare and contrast the two platforms and describe the advantages of each for deploying, managing and monitoring highly scalable applications.

1. © 2015 IBM Corporation1
Java vs Node.js
For Enterprise Web Applications
Chris Bailey
IBM Runtime Technologies

2. © 2015 IBM Corporation2
What Languages do you use?
Java JavaScript Both
0
20
40
60
80
100
120
PercentageofAudience

3. © 2015 IBM Corporation3
What Runtimes do you use?
Server Java Applets JS in Browser Node.js Rhino Nashorn Avatar.js
0
20
40
60
80
100
120
PercentageofAudience

4. © 2015 IBM Corporation4
4
Chris Bailey
IBM Runtime Monitoring and Diagnostics Architect
–14 years working with Java and JVM technologies
–14 months working with Node.js and V8
Current Role(s):
–IBM Java monitoring and diagnostics
–Node Application Metrics project lead
Node Foundation Member
Benchmarking and Performance WGs
JavaOne RockStar
Author on performance and memory analysis
baileyc@uk.ibm.com
chrisbaileyibm
cnbailey
@seabaylea
@Chris__Bailey

5. © 2015 IBM Corporation5

Language Adoption

Deployment Modes

Code Development

Scalability

WebApplication Performance

Under the Hood

Enterprise Architectures
Agenda

6. © 2015 IBM Corporation6
6
Language Adoption

7. © 2015 IBM Corporation7
GitHub Adoption: Java

8. © 2015 IBM Corporation8
GitHub Adoption: JavaScript

9. © 2015 IBM Corporation9
modulecounts.com

10. © 2015 IBM Corporation10
StackOverflow User Survey

11. © 2015 IBM Corporation11
Ratings based on the number of skilled engineers, courses and third party vendors.
Tiobe Community Programming Index

12. © 2015 IBM Corporation12
Indeed.com Job Trends: Java

13. © 2015 IBM Corporation13
Indeed.com Job Trends: JavaScript

14. © 2015 IBM Corporation14
Indeed.com Job Trends
Java
JavaScript

15. © 2015 IBM Corporation15

JavaScript has a large developer base
- #1 on GitHub with 45% more active repositories than Java
- #1 on modulecounts.com with 29% more NPM modules than
Maven
- #1 used language by StackOverflow survey responders
- #6 language on the Tiobe index

Java remains huge, particularly in the enterprise and on the server
- #2 on GitHub with 52% more active repositories than the next
language
- #3 on modulecounts with 73.8% more modules than the next
language
- #2 language on the Tiobe index
- #1 on indeed.com for developer jobs
Language Adoption

16. © 2015 IBM Corporation16
Deployment Modes

17. © 2015 IBM Corporation17

JavaScript is ubiquitous in the browser
- Supported in every browser
- Integration with HTML and CSS

JavaScript is not affected by negative publicity....
Unless it is absolutely necessary to run Java in web browsers, disable it as described
below, even after updating to 7u11. This will help mitigate other Java vulnerabilities that
may be discovered in the future.
This and previous Java vulnerabilities have been widely targeted by attackers, and
new Java vulnerabilities are likely to be discovered. To defend against this and future
Java vulnerabilities, consider disabling Java in web browsers…
Usage in the browser

18. © 2015 IBM Corporation18

Java has a long history on the server
- JPE launched in 1998

Java has rich platform support:
- Linux x86, Linux POWER, zLinux
- Windows, Mac OS, Solaris, AIX, z/OS

JavaScript is a nascent language on the server
- Limited platform support – although its growing
- No API support to interact with the OS

Part of the browser security model
- Frameworks like Node.js have changed that.
Usage on the server

19. © 2015 IBM Corporation19

Single Threaded Event based JavaScript framework
– Uses non-blocking asynchronous I/O

Wraps the Chrome V8 JavaScript engine with I/O interfaces
– Libuv provides interaction with OS/system

Designed to build scalable network applications
– Suited for real time delivery of data to distributed client

Available on a wide set of platforms:
- Linux on x86, ARM, Power and Z
- Windows, Mac OS, Solaris, SmartOS and AIX
libuvV8
Node Bindings
Node Standard Library
C
JavaScript
Server Side JavaScript: Node.js

20. © 2015 IBM Corporation20
Anatomy of a Node.js application
1. package.json
–
Tracks properties, dependencies, version info, commands, etc.
2. JavaScript source files (i.e. app.js)
•To install and execute:
> npm install
> node app.js
20

21. © 2015 IBM Corporation21
Code Development

22. © 2015 IBM Corporation22
var cluster = require('cluster');
var cpus = require('os').cpus().length;
var http = require('http');
if (cluster.isMaster) {
for (var i = 0; i < cpus; i++) {
cluster.fork();
}
cluster.on('death', function(worker) {
console.log("Worker" + worker.pid + "died");
});
} else {
http.createServer(function(request, response) {
response.writeHead(200, {"Content-Type": "text/plain"});
response.write("Hello World!n");
response.end();
}).listen(8080);
}
HTTP Server Example

23. © 2015 IBM Corporation23
var cluster = require('cluster');
var cpus = require('os').cpus().length;
var http = require('http');
if (cluster.isMaster) {
for (var i = 0; i < cpus; i++) {
cluster.fork();
}
cluster.on('death', function(worker) {
console.log("Worker" + worker.pid + "died");
});
} else {
http.createServer(function(request, response) {
response.writeHead(200, {"Content-Type": "text/plain"});
response.write("Hello World!n");
response.end();
}).listen(8080);
}
Clustered HTTP Server Example

24. © 2015 IBM Corporation24
Writing Node.js Code
CodeVolumeRelativetoJava
* based on benchmarksgame benchmarks
Node.js
Development Effort
(lower is better)
- 3x
*or other transactional service

25. © 2015 IBM Corporation25
Writing Node.js Code
regex-dna Spectral-norm fannkuch-redux fasta k-nucleotide Binary-trees n-body Reverse-complement
CodeVolumeRelativetoJava
* based on benchmarksgame benchmarks
Node.js
Development Effort
(lower is better)
- 3x
Avg 1/3rd
less code
*or other transactional service

26. © 2015 IBM Corporation26
Writing Node.js Code
regex-dna Spectral-norm fannkuch-redux fasta k-nucleotide Binary-trees n-body Reverse-complement
CodeVolumeRelativetoJava
* based on benchmarksgame benchmarks
Node.js
Development Effort
(lower is better)
- 3x
●
Node.js has higher developer productivity
Many applications developed with significantly less code
●
Rich module system simplifies development
Reduces need to develop custom code
Avg 1/3rd
less code
*or other transactional service

27. © 2015 IBM Corporation27
Scalability

28. © 2015 IBM Corporation28

One thread (or process) per connection
- Each thread waits on a response
- Scalability determined by the number
of threads

Each thread:
- consumes memory
- is relatively idle

Number of concurrent customers
determined by number of depot workers

Additional customers wait in a queue with no
response
Typical approach to I/O

29. © 2015 IBM Corporation29

One thread multiplexes for multiple
requests
- No waiting for a response
- Handles return from I/O when
notified

Scalability determined by:
- CPU usage
- “Back end” responsiveness

Number of concurrent customers
determined by how fast the food Server
can work

Or until the kitchen gets slammed
Asycnhronous Non-Blocking I/O

30. © 2015 IBM Corporation30

JavaScript is already event based in the browser
- eg. onClick and onMouseOver events

First class functions and closures fit well with events
- Easy to create and pass function callbacks
- Easy to execute callbacks in the context of the event

Node.js execution is based on an event loop
- Asynchronous I/O built in from the ground up

Node.js execution uses a single thread
- No need to worry about locking or shared data
- Most machines are now multi-CPU, so cluster capabilities are
provided
JavaScript and Asynchronous I/O

31. © 2015 IBM Corporation31
var http = require('http');
var server = http.createServer();
server.listen(8080);
server.on('request', function(request, response) {
response.writeHead(200, {"Content-Type": "text/plain"});
response.write("Hello World!n");
response.end();
});
server.on('connection', function(socket) {});
server.on('close', function() {});
server.on('connect', function(socket) {});
server.on('upgrade', function(request, socket, head) {});
server.on('clientError', function(exception, socket) {});
Event based programming

32. © 2015 IBM Corporation32
WebApp Performance

33. © 2015 IBM Corporation33

JSON serialization of a
newly instantiated object

Maps
- Key of message
- Value of Hello,
World!

Example response:
Results from TechEmpower.com Round 9 tests (2014-05-01)
JSON Serialization

34. © 2015 IBM Corporation34

JSON serialization of a
newly instantiated object

Maps
- Key of message
- Value of Hello,
World!

Example response:
Results from TechEmpower.com Round 9 tests (2014-05-01)
JSON Serialization
JavaScript
Java

35. © 2015 IBM Corporation35
JavaScript WebApp Performance
-100
-80
-60
-40
-20
0
20
40
-75
Node.js Performance
(Compared to Java)
JSON Serialization
%ageofJavaPerformance

36. © 2015 IBM Corporation36

Fetches single row from
simple database table

Row serialized as JSON

Example response:
Results from TechEmpower.com Round 9 tests (2014-05-01)
Single Query

37. © 2015 IBM Corporation37

Fetches single row from
simple database table

Row serialized as JSON

Example response:
Results from TechEmpower.com Round 9 tests (2014-05-01)
Single Query
JavaScript
Java

38. © 2015 IBM Corporation38
JavaScript WebApp Performance
-100
-80
-60
-40
-20
0
20
40
-60.5
Node.js Performance
(Compared to Java)
JSON Serialization
Single Query
%ageofJavaPerformance

39. © 2015 IBM Corporation39

Fetches multiple rows from
a simple database table

Rows serialized as JSON

Example response:
Results from TechEmpower.com Round 9 tests (2014-05-01)
Multiple Queries

40. © 2015 IBM Corporation40

Fetches multiple rows from
a simple database table

Rows serialized as JSON

Example response:
Results from TechEmpower.com Round 9 tests (2014-05-01)
Multiple Queries
JavaScript
Java

41. © 2015 IBM Corporation41
JavaScript WebApp Performance
-100
-80
-60
-40
-20
0
20
40
-18
Node.js Performance
(Compared to Java)
JSON Serialization
Single Query
Multiple Queries
%ageofJavaPerformance

42. © 2015 IBM Corporation42

Fetches multiple rows from
a simple database table

Converts rows to objects
and modifies one attribute
of each object

Updates each associated
row and serializes as
JSON

Example Response:
Data Updates
Results from TechEmpower.com Round 9 tests (2014-05-01)

43. © 2015 IBM Corporation43

Fetches multiple rows from
a simple database table

Converts rows to objects
and modifies one attribute
of each object

Updates each associated
row and serializes as
JSON

Example Response:
Data Updates
Results from TechEmpower.com Round 9 tests (2014-05-01)
JavaScript
Java

44. © 2015 IBM Corporation44
JavaScript WebApp Performance
-100
-80
-60
-40
-20
0
20
40
28
Node.js Performance
(Compared to Java)
JSON Serialization
Single Query
Multiple Queries
Data Updates
%ageofJavaPerformance

45. © 2015 IBM Corporation45

Computation speed is (much) slower than Java

I/O speed is higher than Java
JavaScript WebApp Performance
-100
-80
-60
-40
-20
0
20
40
-75 -60.5 -18
28
Node.js Performance
(Compared to Java)
JSON Serialization
Single Query
Multiple Queries
Data Updates
%ageofJavaPerformance
More
Computation
More
I/O

46. © 2015 IBM Corporation46
JavaScript
on the JVM?

47. © 2015 IBM Corporation47
Nashorn and Avatar.js

Nashorn JavaScript engine delivered in JDK8
– Utilizes new JVM level features
for performance

Avatar.js provides Node.js support on Nashorn

48. © 2015 IBM Corporation48
Nashorn and Avatar.js

Nashorn JavaScript engine delivered in JDK8
– Utilizes new JVM level features
for performance

Avatar.js provides Node.js support on Nashorn

Results of “Octane” JavaScript benchmark*:
– Node.js is 4.8x faster
– Avatar.js is >10x larger
* Using Java 8 pre-u20

49. © 2015 IBM Corporation49
Nashorn and Avatar.js

Nashorn JavaScript engine delivered in JDK8
– Utilizes new JVM level features
for performance

Avatar.js provides Node.js support on Nashorn

Results of “Octane” JavaScript benchmark*:
– Node.js is 4.8x faster
– Avatar.js is >10x larger
* Using Java 8 pre-u40
Feb 12th
, 2015: Avatar is “put on hold”
https://blogs.oracle.com/theaquarium/entry/project_avatar_update

50. © 2015 IBM Corporation50
JavaScript Performance
Under the Hood

51. © 2015 IBM Corporation51

Dynamic nature of JavaScript makes JIT optimizations more difficult
– Any variable could be a function or data
– Variables must be tested to determine how to handle handle them
– Variables can change, so what worked previously many not apply in future

Example: the use of '+':
• number + number → addition
• string involved? → concatenation
• objects involved? → convert to primitives then addition or
concatenation
• Functions involved? → ?
JIT Compilation

52. © 2015 IBM Corporation52
Execution flow for '+' operator
int a
int b
int c = a + b
Addition
Return Result
Java

53. © 2015 IBM Corporation53
Execution flow for '+' operator
int a
int b
int c = a + b
Addition
Return Result
var a
var b
var c = a + b
First Time?
NO
Use Previous
Types
Worked?
Return Result
YES
Type of A?
NO
YES
Type of B?String Concatenate
String
or
Number
Number Type of B? Number Addition
UndefinedString
Java JavaScript

54. © 2015 IBM Corporation54

Dynamically typed languages are harder to manage under the hood

They subsequently have lower runtime performance for computational tasks

Highlighted in TechEmpower benchmarks:
Dynamically vs Statically Types Languages
-70
-60
-50
-40
-30
-20
-10
0
Dynamic vs Statically Typed Language Performance
JSON
Single
Multi
Updates
Bestdynamiccomparedtobeststaticasbaseline

55. © 2015 IBM Corporation55
Writing
Node.js Code

56. © 2015 IBM Corporation56
Not So Easy: JavaScript Syntax Fun
> '5' – 3
2 // Weak typing, implicit conversion
> '5' + 3
'53' // ...but not consistent
> '5' – '4'
1 // String minus String = Integer??
> '5' + + '4'
54 // Multiple +'s are ok
> 'Hello' + 'World'
'HelloWorld' // Ok, that's expected
> 'Hello' + + 'World'
'HelloNaN' // ...but that isn't

57. © 2015 IBM Corporation57
Not So Easy: JavaScript Syntax Fun
> '5' – 3
2 // Weak typing, implicit conversion
> '5' + 3
'53' // ...but not consistent
> '5' – '4'
1 // String minus String = Integer??
> '5' + + '4'
54 // Multiple +'s are ok
> 'Hello' + 'World'
'HelloWorld' // Ok, that's expected
> 'Hello' + + 'World'
'HelloNaN' // ...but that isn't

58. © 2015 IBM Corporation58
Not So Easy: JavaScript Syntax Fun
> '5' – 3
2 // Weak typing, implicit conversion
> '5' + 3
'53' // ...but not consistent
> '5' – '4'
1 // String minus String = Integer??
> '5' + + '4'
54 // Multiple +'s are ok
> 'Hello' + 'World'
'HelloWorld' // Ok, that's expected
> 'Hello' + + 'World'
'HelloNaN' // ...but that isn't

59. © 2015 IBM Corporation59
Not So Easy: JavaScript Syntax Fun
> '5' – 3
2 // Weak typing, implicit conversion
> '5' + 3
'53' // ...but not consistent
> '5' – '4'
1 // String minus String = Integer??
> '5' + + '4'
54 // Multiple +'s are ok
> 'Hello' + 'World'
'HelloWorld' // Ok, that's expected
> 'Hello' + + 'World'
'HelloNaN' // ...but that isn't

60. © 2015 IBM Corporation60
Not So Easy: JavaScript Syntax Fun
> '5' – 3
2 // Weak typing, implicit conversion
> '5' + 3
'53' // ...but not consistent
> '5' – '4'
1 // String minus String = Integer??
> '5' + + '4'
54 // Multiple +'s are ok
> 'Hello' + 'World'
'HelloWorld' // Ok, that's expected
> 'Hello' + + 'World'
'HelloNaN' // ...but that isn't

61. © 2015 IBM Corporation61
Not So Easy: JavaScript Syntax Fun
> '5' – 3
2 // Weak typing, implicit conversion
> '5' + 3
'53' // ...but not consistent
> '5' – '4'
1 // String minus String = Integer??
> '5' + + '4'
54 // Multiple +'s are ok
> 'Hello' + 'World'
'HelloWorld' // Ok, that's expected
> 'Hello' + + 'World'
'HelloNaN' // ...but that isn't

62. © 2015 IBM Corporation62
Not So Easy: JavaScript Syntax Fun
> '5' – 3
2 // Weak typing, implicit conversion
> '5' + 3
'53' // ...but not consistent
> '5' – '4'
1 // String minus String = Integer??
> '5' + + '4'
54 // Multiple +'s are ok
> 'Hello' + 'World'
'HelloWorld' // Ok, that's expected
> 'Hello' + + 'World'
'HelloNaN' // ...but that isn't

63. © 2015 IBM Corporation63
Not So Easy: JavaScript Syntax Fun
> '5' – 3
2 // Weak typing, implicit conversion
> '5' + 3
'53' // ...but not consistent
> '5' – '4'
1 // String minus String = Integer??
> '5' + + '4'
54 // Multiple +'s are ok
> 'Hello' + 'World'
'HelloWorld' // Ok, that's expected
> 'Hello' + + 'World'
'HelloNaN' // ...but that isn't

64. © 2015 IBM Corporation64
Not So Easy: JavaScript Syntax Fun
> '5' – 3
2 // Weak typing, implicit conversion
> '5' + 3
'53' // ...but not consistent
> '5' – '4'
1 // String minus String = Integer??
> '5' + + '4'
54 // Multiple +'s are ok
> 'Hello' + 'World'
'HelloWorld' // Ok, that's expected
> 'Hello' + + 'World'
'HelloNaN' // ...but that isn't

65. © 2015 IBM Corporation65
Not So Easy: JavaScript Syntax Fun
> '5' – 3
2 // Weak typing, implicit conversion
> '5' + 3
'53' // ...but not consistent
> '5' – '4'
1 // String minus String = Integer??
> '5' + + '4'
54 // Multiple +'s are ok
> 'Hello' + 'World'
'HelloWorld' // Ok, that's expected
> 'Hello' + + 'World'
'HelloNaN' // ...but that isn't

66. © 2015 IBM Corporation66
Not So Easy: JavaScript Syntax Fun
> '5' – 3
2 // Weak typing, implicit conversion
> '5' + 3
'53' // ...but not consistent
> '5' – '4'
1 // String minus String = Integer??
> '5' + + '4'
54 // Multiple +'s are ok
> 'Hello' + 'World'
'HelloWorld' // Ok, that's expected
> 'Hello' + + 'World'
'HelloNaN' // ...but that isn't

67. © 2015 IBM Corporation67
Not So Easy: JavaScript Syntax Fun
> '5' – 3
2 // Weak typing, implicit conversion
> '5' + 3
'53' // ...but not consistent
> '5' – '4'
1 // String minus String = Integer??
> '5' + + '4'
54 // Multiple +'s are ok
> 'Hello' + 'World'
'HelloWorld' // Ok, that's expected
> 'Hello' + + 'World'
'HelloNaN' // ...but that isn't

68. © 2015 IBM Corporation68
Not So Easy: JavaScript Syntax Fun
> '5' + - '5'
'5-2' // I can just about see that works
> var x = 3
undefined
> '5' – x + x
5 // Ok, that makes sense
> var x = 3
undefined
> '5' + x - x
50 // What???

69. © 2015 IBM Corporation69
Not So Easy: JavaScript Syntax Fun
> '5' + - '5'
'5-5' // I can just about see that works
> var x = 3
undefined
> '5' – x + x
5 // Ok, that makes sense
> var x = 3
undefined
> '5' + x - x
50 // What???

70. © 2015 IBM Corporation70
Not So Easy: JavaScript Syntax Fun
> '5' + - '5'
'5-2' // I can just about see that works
> var x = 3
undefined
> '5' – x + x
5 // Ok, that makes sense
> var x = 3
undefined
> '5' + x - x
50 // What???

71. © 2015 IBM Corporation71
Not So Easy: JavaScript Syntax Fun
> '5' + - '5'
'5-2' // I can just about see that works
> var x = 3
undefined
> '5' – x + x
5 // Ok, that makes sense
> var x = 3
undefined
> '5' + x - x
50 // What???

72. © 2015 IBM Corporation72
Not So Easy: JavaScript Syntax Fun
> '5' + - '5'
'5-2' // I can just about see that works
> var x = 3
undefined
> '5' – x + x
5 // Ok, that makes sense
> var x = 3
undefined
> '5' + x - x
50 // What???

73. © 2015 IBM Corporation73
Not So Easy: JavaScript Syntax Fun
> '5' + - '5'
'5-2' // I can just about see that works
> var x = 3
undefined
> '5' – x + x
5 // Ok, that makes sense
> var x = 3
undefined
> '5' + x - x
50 // What???

74. © 2015 IBM Corporation74
(Micro)Service
Topologies

75. © 2015 IBM Corporation75
Enterprise Application Platform – Web Applications
Operations and Management
Admin Analytics
LoadBalancer
LoadBalancer
HTTP
Monitoring ScalingAnalytics Diagnostics

76. © 2015 IBM Corporation76
Questions

77. © 2015 IBM Corporation77