Slides for the Introduction to hazelcast talk which I've given for Java User Group Graz at November 2014.

1. introduction to
1
emin demirci

2. introduction
distributed computing
in-memory data grids
hazelcast
code samples
demo
internals
q/a
2
agenda

3. about me
3
core developer at hazelcast
holds bsc. computer engineering
started programming some time ago,
then turned into a career
lives in beautiful istanbul
interested in distributed systems

4. distributed computing
4
use of bunch of computers to solve a
computational problem
problem is divided into multiple tasks
and they are solved by one or more
computers
computers communicate each other
by sending messages

5. in-memory data grids
5
middleware software
shared nothing architecture
manages objects across distributed
servers in the RAM
ability to scale
provides fault tolerance

6. why use an imdg?
6
performance - ram is faster
flexibility - rich set of data structures
operations - easy to scale/maintain

7. other imdg solutions
7
oracle coherence
ibm extremescale
vmware gemfire
gigaspaces
redhat infinispan
gridgain
terracotta

8. 8
what is ?

9. an open-source project
9

10. a company
10

11. a company
11
hazelcast enterprise edition
management center
enterprise support
training / consulting
offices in istanbul (r&d), palo alto(hq)
and london

12. an open-source project
12
leading open-source in-memory data
grid.
dead simple distributed
programming
easy way to scale applications
simple api
built with in Istanbul

13. use cases
13
scaling your application
sharing data across cluster
partitioning data
sending/receiving messages
load balancing
session replication
parallel task processing on multiple
machines
…

14. how differs ?
14
apache licensed open source
lightweight w/o any dependency
ease of use and more fun !

15. 15
fact : every ~0.4 second a hazelcast node is started
around the world
a lot of developers :)
who uses ?

16. who uses ?
16

17. 17
what is ?
distributed impl. of Java Collections
dynamic clustering, backup and
failover
transaction support ( two phase, XA)
distributed execution framework
map/reduce api
distributed queries
native Java, C#, C++ clients

18. starting a hazelcast instance
18

19. hazelcast instance api
19

20. quick demo
20

21. data partitioning
21
Drawing by Benjamin Erb http://berb.github.io/diploma-thesis/original/resources/cons_hash.svg

22. distributed map
22

23. distributed queries
23

24. processing entries
24

25. entry processors
25

26. distributed queue
26

27. distributed executor service
27

28. distributed lock
28

29. aggregations
29

30. wan replication
30
active/active
active/passive

31. wan replication
31
• High availability

32. hazelcast architecture
32

33. service provider interface
( SPI )
33
roll your own services
extend hazelcast based on your needs !
hierarchical lock service
priority queue
scheduled executor service
distributed actors
anything you can think of !
check out SPI section of the hazelcast
documentation

34. session replication
34
servlet filter based
just put hazelcast filter to your web.xml
native tomcat/jetty plugins (enterprise)

35. amazon ec2 support
35

36. near cache
36
client side
node side

37. map/reduce
37

38. map/reduce
38

39. 39
hazelcast stabilizer

40. hazelcast stabilizer
40

41. 41
hazelcast stabilizer

42. Thank you ! :)
emin@hazelcast.com
42
any questions ?
we are hiring, check out
hazelcast.com/careers