kai wähner – real world use cases for realtime in-memory computing - nosql matters barcelona 2014

26
In-Memory Computing “Real World Use Cases” Kai Wähner Technical Lead [email protected] @KaiWaehner www.kai-waehner.de LinkedIn / Xing Please connect!

Upload: nosqlmatters

Post on 12-Jul-2015

384 views

Category:

Data & Analytics


0 download

TRANSCRIPT

In-Memory Computing  “Real  World  Use  Cases”  

Kai WähnerTechnical [email protected]@KaiWaehnerwww.kai-waehner.deLinkedIn / Xing à Please connect!

Kai Wähner

Consulting Developing Coaching Speaking

Writing Selling

Main Tasks Requirements Engineering

Enterprise Architecture Management Business Process Management

Architecture and Development of Applications Service-oriented Architecture

Integration of Legacy Applications Cloud Computing

Big Data

Contact Email: [email protected] Blog: www.kai-waehner.de/blog

Twitter: @KaiWaehner Social Networks: LinkedIn, Xing

Disclaimer

!    

These  opinions  are  my  own  and  do  not  necessarily  represent  my  employer  

Key Messages

In-Memory Computing is used for Acting in Real-Time!

In-Memory Computing is NOT just Caching!

Eventing and Fault-Tolerance move In-Memory to another Level!

© Copyright 2000-2014 TIBCO Software Inc. 5  

Agenda

•  Introduction to In-Memory Computing•  Use Cases / Customer Success Stories

© Copyright 2000-2014 TIBCO Software Inc. 6  

Agenda

•  Introduction to In-Memory Computing•  Use Cases / Customer Success Stories

Time  

Business Value

Business Event

Data Ready for Analysis

Analysis Completed

Decision Made

$$$$  

$$$  

$$  

$   Action Taken

Business Value of Events over Time

In-Memory Computing and Event Processing

speeds action and increases business

value by seizing opportunities while

they matter

© Copyright 2000-2014 TIBCO Software Inc. 8  

•  Hardware costs declining•  Data Processing Requirements

exploding•  Traditional Approaches not

scaling–  Relational Databases–  Clustered Databases–  In-Memory Caches–  Messaging Systems

Drivers for In-Memory Computing

© Copyright 2000-2014 TIBCO Software Inc. 9  

Database Landscape in 2014

h9p://blogs.the451group.com/  informaCon_management/2014/03/18/  updated-­‐data-­‐plaIorms-­‐landscape-­‐  

map-­‐february-­‐2014/  

© Copyright 2000-2014 TIBCO Software Inc. 10  

Agenda

•  Introduction to In-Memory Computing•  Use Cases / Customer Success Stories

LOADER  

   

Caching for Fast Data Access

•  Cache  to  slower  systems  •  Read-­‐only  •  Not  the  system  of  record    •  No  persistence  required  •  Side  benefit:  Backend  load  

is  reduced  

   

SERVICE  

Caching + Dynamic Load

•  Dynamically  loaded  into  Memory  when  the  data  is  first  accessed  by  a  client  applicaCon  

•  Service  can  present  a  standard  interface    

•  Client  applicaCons  are  not  required  to  implement  any  In-­‐Memory  specific  code  

(1)  Check  Cache  

(2)  Load  from  DB  if  not  in  Cache  

Routing Messages to Back-Office Applications

•  Receive  a  common  data  feed  that  needs  to  be  parsed  and  routed  to  several  back-­‐office  applicaCons  can  use    

•  In-­‐Memory  holding  reference  informaCon  for  the  rouCng  applicaCon.  The  router  can  quickly  determine  where  to  send  the  data.    

•  Examples:  Bank  payments,  insurance  claims  processing  

Off-loading expensive systems

Expensive  in  terms  of  response  Cme  and  /  or  transacCon  costs!  

Personalized Customer Experience

“With  38  million  fans,  MGM  knows  how  to  put  its  customers  first,  it  takes  more  than  a  smile  too.  Customers  want  a  personalized,  tailored  experience,  one  that  knows  their  name  and  can  anCcipate  their  needs.  With  the  help  of  TIBCO  technologies  that  leverage  big  data  and  give  customers  a  digital  idenCty,  MGM  can  send  personalized  offers  directly  to  customers,  save  them  a  seat,  and  have  their  favorite  drink  on  the  way.  With  mulCple  customer  touch  points  and  channels,  MGM  can  reach  customers  in  more  ways,  and  in  more  places,  than  ever  before.”    

h9ps://www.youtube.com/watch?v=X-­‐7S3kCOx9k  

Latency  Problems:  •  Several  Legacy  Systems  •  Processing  via  ERP,  CRM,  Host,  etc.    In-­‐Memory:  •  Enable  Real  Time  •  Only  customers  that  have  checked  in  •  System  of  Record  

Handling temporary spikes on a slow ‘system of record’

•  An  In-­‐Memory  event  listener  gets  noCfied  whenever  a  data  value  is  changed  and  sends  updates  through  a  message  queue  for  updaCng  the  master  system  of  record.  

•  The  back  office  system  can  also  be  updated  through  other  channels.  •  Examples:  Christmas  Shopping  in  E-­‐Commerce,  Ticket  Sales,  Online  Bekng  

à  In-­‐Memory  as  “system  of  record”    à  OpConal:  PersisCng  data  on  the  local  file  system  (rather  than  requiring  a  database  for  persisCng  data    

Operational Data Store (Local File System)

•  Low-­‐latency,  high-­‐throughput  operaConal  data  –  Customer  data:  e.g.  account  status  and  balance,  

purchase  history:  real-­‐Cme  loyalty  (promoCons,    cross-­‐selling),  fraud  detecCon,  ...  

–  Market  data:  e.g.  risk  assessment,  porIolio  mgmt,  producCon  output  opCmizaCon,  buyer-­‐seller  matching  

–  Sensor  data:  e.g.  smart  metering  /  grid,  public  transport  safety  –  Track  and  trace:  e.g.  barcode  scans,  RFID:  logisCcs,  airlines  

•  Why  In-­‐Memory?  –  Much  faster  than  tradiConal  DB,  especially  many  small  transacCons  (XTP)  –  State  /  data  management  not  addressed  by  messaging  soluCons  –  EvenCng  is  a  first  class  feature,  changes  can  be  ‘pushed’  in  real-­‐Cme  to  interested  parCes  

(subscribe  to  changes,  conCnuous  queries)  –  Provides  for  distributed  process  synchronizaCon  –  Integrated  with  CEP  engine  (e.g.  TIBCO  BusinessEvents)  

Operational Data Store (Local File System)

Situation•  Master data management system stores over 800 million customer records across more than 30 enterprise apps. •  Stores real-time inventory data to enable ‘Buy online and pick-up at store’ and ‘Smart fulfillment’ features

Problem •  Due to lack of correlation between Point of Sale data and inventory, the website contained outdated inventory data.

Products were listed as out of stock when there was actually inventory. •  Need to leverage store inventory as well as inventory located fulfillment centers

Solution•  In-Memory stores real-time inventory data for the website, the fulfillment application, and other applications that need

access to inventory dataBusiness Impact

•  Reduction in customer churn•  Intelligent fulfillments leading to greater customer satisfaction•  Improved overall efficiency of fulfillment centers and store inventory

Retailer: Inventory Management

Distribution of Rapidly Changing Data

à   Examples  are  monitoring  data  for  a  power  plant,  stock  market  data,  telemetry  data  for  a  complex  system  (example,  a  satellite),  or  the  status  and  locaCon  of  packages  for  a  major  logisCcs  or  shipping  company.    

Telco: Real-Time Offer Generation and Fulfillment by Different Subcontractors

ReloadGive 100 free SMS to subscriber who tops-up

Total: 12 mio top-up / dayPeak: 300 top-up per sec

Purchase 3G PackageCross-sell Voice/SMS package to subscriber who purchases 3G Mobile Package

Total: 3 mio / dayPeak: 50 events per sec

Voice CallGive discount VOIP package to subscriber who makes a IDD call

Total: 200 mio / dayPeak: 12,000 events per sec

SMS UsageGive discounted SMS package to subscriber who sends SMS more than 10 times a day

Total: 750 mio / dayPeak: 27,000 events per sec

Event Cloud

Purchase BB Package

Reload

Voice Call

IDD Call

OnNet Call

SMS Usage

Event Handling and Processing Touchpoint Integration Billing, Offer

Fulfilled

Fulfill SMS Package

Fulfill 3G Package

Fulfill Voice Package

Fulfill SMS Package

46.7 million subscribers 2,000 SMS notifications per

seconds

500 offer fulfillments per

second

Offer MessageReminderMessage

FulfillmentMessage

State-­‐full  Data  

Storing State-full Data for Enterprise Applications

Super Fast Compute Grid for Intermediary Calculations for Analytics

•  Technical  issues  in  distributed  grid  compuCng  with  large  scale  data  –  Work  load  distribuCon  –  Process  synchronizaCon  –  Data  transfer  

•  Examples  –  Risk  assessment  and  management  –  OpCmizaCon  problems:  scheduling,  cargo  assignment,  load  distribuCon  in  

power  network  /  grid  

•  Why  In-­‐Memory?  –  Many  useful  synchronizaCon  features  (e.g.  atomic  “take”)  –  LocaCon  transparency  and  fault-­‐tolerance  –  Real-­‐Cme  instead  of  nightly  /  weekly  /  ...  Data-­‐Warehousing  approach  

Super Fast Compute Grid for Intermediary Calculations for Analytics

Key Messages

In-Memory Computing is used for Acting in Real-Time!

In-Memory Computing is NOT just Caching!

Eventing and Fault-Tolerance move In-Memory to another Level!

Questions? Kai Wähner [email protected] @KaiWaehner www.kai-waehner.de LinkedIn / Xing à Please connect!