30
A universal Data & State sharing Fabric March 2013 Shay Hassidim Deputy CTO 1

Telecom universal datastatesharingfabric

Embed Size (px)

DESCRIPTION

 

Citation preview

Page 1: Telecom universal datastatesharingfabric

A universal Data & State sharing FabricMarch 2013

Shay HassidimDeputy CTO

1

Page 2: Telecom universal datastatesharingfabric

2

Agenda

• Company Snapshot• Shared Data & State Fabric

– Problem Statement – Existing Solutions – Pros & Cons– IMDG – Fundamentals & Case Study

Page 3: Telecom universal datastatesharingfabric

GigaSpaces – Telco Customers and Partners

3

Founded 2000 , 200 Customers , Sales: US , EU , APAC , RND: TLV , Main Investor: Intel

Page 4: Telecom universal datastatesharingfabric

4

Problem Description• Telecom systems internal systems currently use a variety of

MOM (AMQP, XMPP) , IM protocols , Network based protocols, Application level protocols (SIP), HTML based protocols (Rest, SOAP) or DB (RDBMS, No SQL) to communicate with each other.– Introduce complexity, cost/ROI concerns, increased time to

market, impact customer satisfaction

• A universal data & state sharing fabric connecting the different products / systems / components is required– Real time response– Total reliability– Scalable & Highly Available

Page 5: Telecom universal datastatesharingfabric

5

Possible Solutions - #1:Database

• Pros– Everybody knows databases. No need to educate the

developers.• Cons

– Users need to reshape the data structure to use a specific model (Table , Document). Lock-in.

– No proactive event based messaging fabric. Requires external messaging integration

– Management overhead– Not designed for real-time communication. Mostly

disk based storage medium.

Page 6: Telecom universal datastatesharingfabric

Possible Solutions - #2:ESB

• Pros– Designed for data transformation and

routing between alien systems – Protocol translation focused

• Cons– Does not maintain state. No state sharing.

Getting the latest state requires a DB/Cache– Not designed for real-time communication

6

Page 7: Telecom universal datastatesharingfabric

7

What if we could combine both DB and ESB but avoid all the

cons?+ =

Page 8: Telecom universal datastatesharingfabric

Enter IMDG• Roots in the Tuple Space concept, RDBMS, OODBMS,

MOM• Pros

– State sharing in real time in Massive scale– Built-in proactive event based fabric

• Fused to the engine core. No external subsystem– Support both Key/Value, SQL Based data access and

Message deliver modes with the same API• Topic , Queue, Pub/Sub

– Support locality of reference• Collocate Data and business logic

• Cons– No full SQL Query support as a RDBMS– Requires some education/ramp up

8

Page 9: Telecom universal datastatesharingfabric

9

GigaSpaces IMDG Basic Operations

Application

Space

Take

Application

Space

Read

Application

Space

WriteMultiple

Application

Space

Write

Application

Space

ReadMultiple

Application

Space

TakeMultiple

Application

Space

Execute

Application

Space

Notify

http://www.gigaspaces.com/docs/JavaDoc9.1/org/openspaces/core/GigaSpace.html

Application

Space

Changeset

putInMapincrement

deccrementaddToCollection

Page 10: Telecom universal datastatesharingfabric

10

POJOJavaSpaceMapJPAJMSJDBC

IMDG Real-Time InteroperabilityDocument

IMDG

Page 11: Telecom universal datastatesharingfabric

Application-IMDG Topologies

ApplicationProcess

IMDG Process

Remote call

Application Process

IMDG Instance

Application

Local call

Remote

Collocated

11

Page 12: Telecom universal datastatesharingfabric

0

50

100

150

200

250

300

5

200

10

300

180

GigaSpaces IMDG Latency BenchmarkM

icro

seco

nds

IMDG Real-Time Low-Latency Performance

http://blog.gigaspaces.com/2010/12/06/possible-impossibility-the-race-to-zero-latency/

Async operationNo Serialization

No Network usage Async operation

12

Benchmark using Cisco UCS , 2.93GHz CPU

Page 13: Telecom universal datastatesharingfabric

13

Cassandra vs. GigaSpaces

1 2 3 4 5 6 7 8 100

10,000,000

20,000,000

30,000,000

40,000,000

50,000,000

0

5,000

10,000

15,000

20,000

25,000

Cassandra vs. GigaSpaces Read Benchmark

# of Client Threads

Gig

aSpa

ces

TPS

Cass

andr

a TP

S

1 2 3 4 5 6 7 8 100

200,000

400,000

600,000

800,000

1,000,000

1,200,000

1,400,000

020,00040,00060,00080,000100,000120,000140,000160,000180,000

Cassandra vs. GigaSpaces Write Benchmark

# of Client Threads

Gig

aSpa

ces

TPS

Cass

andr

a TP

S

GigaSpaces is 1000-2000 times faster

GigaSpaces is 3-7 times faster

Benchmark using Cisco UCS , 2.93GHz CPU

Page 14: Telecom universal datastatesharingfabric

ESB Performance

14

http://esbperformance.org/display/comparison/ESB+Performance - August 2012

2 orders of magnitude slower than IMDG

Page 15: Telecom universal datastatesharingfabric

15

IMDG Use Cases

• Database Performance, Scaling • State replication for correlation• Distributed Stateful Processing• State replication over the WAN

Page 16: Telecom universal datastatesharingfabric

• Description– Device activity history – Last n activities

• Challenge– Central DB bottleneck

• Desired solution– Use a distributed cache to front end the database– Preload the data from the DB– Keep the DB in-synch

• Requirements– High consistency– High Performance (<10,000 TX/Sec)

Database Performance, Scaling

16

Page 17: Telecom universal datastatesharingfabric

• Use IMDG to front-end the database

• Keep the DB in-synch– Load the data into memory

using Object based model– Write behind all updates

• Use the IMDG as the system of record– Rich Query (SQL,..)– Transaction support

Database Performance, Scaling Solution Architecture

17

Application

Database 1

load

Database 2 Database 3

Write behind

Page 18: Telecom universal datastatesharingfabric

• Description– State correlation

• Challenge– High availability

• Desired solution– Peer model , multi-master architecture

• Requirements– CAP based consistency

State Replication for correlation

18

Page 19: Telecom universal datastatesharingfabric

• Store correlation state in the IMDG

• Can use replicated or partitioned IMDG topology

• Can use Client-side cache for greater performance optimization of read operation

State Replication for correlationSolution Architecture

19

Application

Application

Application

Page 20: Telecom universal datastatesharingfabric

• Description– Workflow of distributed process– 5-10 steps per process

• Challenge– State information is too big to transfer

• Desired solution– Shared consistent state

• Requirements– Scale to deal with large amount of nodes (>1000)

Distributed Stateful Processing

20

Page 21: Telecom universal datastatesharingfabric

• Store workflow state in the IMDG

• Use pub/sub model to synchronize the state

• Use Template matching / SQL for querying the state of a particular object in the IMDG

Distributed Stateful ProcessingSolution Architecture

POCESS A

STATE A PRCESS BSTATE

BPROCESS

CSTATE

C

21

Page 22: Telecom universal datastatesharingfabric

22

Sharing Device State Across Different DC

• Description– Systems/devices running across different data centers– Need to share state in real time

• Challenge– Requires reliable and scalable data replication over the WAN

• Desired solution– Simple secured point2point state replication over the WAN

• Requirements– Support low-bandwidth , high-latency networks

Page 23: Telecom universal datastatesharingfabric

Replication over the WAN SupportSolution Architecture

London

Hong Kong

New York

23

Page 24: Telecom universal datastatesharingfabric

GigaSpaces IMDG WAN Replication Features

24

Any Replication Topology

High Availability

Data Filtering

Cloud Enabled

Fully Transactional

Custom Conflict Resolution

Interoperable

Single Click Bootstrapping

Security Optimized Connection

Page 25: Telecom universal datastatesharingfabric

XAP – One Product for In Memory Computing

Java-.NET-C++Spring, JPA,JMS

JDBC, MapSchema-Free

One Model for all components

-Clustering -Security

-HA -OA&M

Virtualize All Middleware Components

Customize Application

Management Rules &

Workflows

Real-Time Automated Deployment Monitoring

Management

25

Page 26: Telecom universal datastatesharingfabric

Consistent Management & Monitoring Module (“Cloudify”)

26

Application recipe uses domain-specific language (DSL) to describe the application life cycle

All necessary plumbing provided out of the box

Configuration and setup separate from the application

recipe

Page 27: Telecom universal datastatesharingfabric

A Typical App…

Page 28: Telecom universal datastatesharingfabric

28

Extensive Platform Support

Page 29: Telecom universal datastatesharingfabric

29

Summary – GigaSpaces IMDG

• Designed for real-time data-driven interoperability

• Management of data-centric architecture made easy

• Supports high-end, complex applications• Blazing fast. Highly-Available.

Page 30: Telecom universal datastatesharingfabric

30

Thank You!www.gigaspaces.com