Modernizing Business with BIG DATA |
| Без темы | ||
| << Model SLO – Algebra I | Modular Origami Cube >> |
Автор: Chuck Jones - Corporate Services. Чтобы познакомиться с картинкой полного размера, нажмите на её эскиз. Чтобы можно было использовать все картинки для урока английского языка, скачайте бесплатно презентацию «Modernizing Business with BIG DATA.ppt» со всеми картинками в zip-архиве размером 3768 КБ.
Modernizing Business with BIG DATA
содержание презентации «Modernizing Business with BIG DATA.ppt»| Сл | Текст | Сл | Текст |
| 1 | Modernizing Business with BIG DATA. | 15 | Page 15. Mainframe batch business process |
| Aashish Chandra Divisional VP, Sears | would not scale Needed to process 100 | ||
| Holdings Global Head, Legacy | times more detail to handle business | ||
| Modernization, MetaScale. | critical functionality Business need | ||
| 2 | Big Data fueling Enterprise Agility. | required processing billions of records | |
| Harvard Business Review refers Sears | from 30 input data sources Complex | ||
| Holdings Hadoop use case - Big Data's | business logic and financial calculations | ||
| Management Revolution! Sears eschews | SLA for this cyclic process was 2 hours | ||
| IBM/Oracle for open source and self build. | per run. Mainframe Scalability: Unable to | ||
| Sears’ Big Data Swap Lesson: Functionality | Scale 100 fold. Data Sources: 30+. | ||
| over price? How banks can benefit from | Mainframe: 100 MIPS on 1% of data. Input | ||
| real-time Big Data analytics? | Records: Billions. | ||
| 3 | Legacy Rides The Elephant. Hadoop has | 16 | The Result – Use-Case #2. Hadoop. |
| changed the enterprise big data game. Are | Business Problem: Mainframe batch business | ||
| you languishing in the past or adopting | process would not scale Needed to process | ||
| outdated trends? | 100 times more detail to handle rollout of | ||
| 4 | Journey to the world with NO | high value business critical functionality | |
| Mainframes.. Cost Savings Open Source | Time sensitive business need required | ||
| Platform Simpler & Easier Code | processing billions of records from 30 | ||
| Business Agility Business & IT | input data sources Complex business logic | ||
| Transformation Modernized Systems IT | and financial calculations SLA for this | ||
| Efficiencies. Mainframe Migration. II. | cyclic process was 2 hours per run. | ||
| Mainframe ONLINE Tool based Conversion | Teradata & Mainframe Data on Hadoop. | ||
| Convert COBOL & JCL to Java. I. | JAVA UDFs for financial calculations. | ||
| Mainframe Optimization 5% ~ 10% MIPS | Scalable Solution in 8 Weeks. Implemented | ||
| Reduction Quick Wins with Low hanging | PIG for Processing. Processing Met Tighter | ||
| fruits. III. Mainframe BATCH ETL | SLA. $600K Annual Savings. 6000 Lines | ||
| Modernization Move Batch Processing to | Reduced to 400 Lines of PIG. Mainframe | ||
| Hadoop. High TCO. Inert Business | Scalability: Unable to Scale 100 fold. | ||
| Practices. Optimize. Convert. PiG / Hadoop | Data Sources: 30+. Mainframe: 100 MIPS on | ||
| Rewrites. Resource Crunch. | 1% of data. Input Records: Billions. | ||
| 5 | Why Hadoop and Why Now? THE | 17 | The Challenge – Use-Case #3. Hadoop. |
| ADVANTAGES: Cost reduction Alleviate | Mainframe unable to meet SLAs on growing | ||
| performance bottlenecks ETL too expensive | data volume. Data Storage: Mainframe DB2 | ||
| and complex Mainframe and Data Warehouse | Tables. Price Data: 500M Records. | ||
| processing ? Hadoop THE CHALLENGE: | Processing Window: 3.5 Hours. Mainframe | ||
| Traditional enterprises lack of awareness | Jobs: 64. | ||
| THE SOLUTION: Leverage the growing support | 18 | The Result – Use-Case #3. Hadoop. | |
| system for Hadoop Make Hadoop the data hub | Business Problem: Mainframe unable to meet | ||
| in the Enterprise Use Hadoop for | SLAs on growing data volume. Source Data | ||
| processing batch and analytic jobs. | in Hadoop. $100K in Annual Savings. | ||
| 6 | The Classic Enterprise Challenge. | Maintenance Improvement – <50 Lines PIG | |
| 7 | The Sears Holdings Approach. 1. 2. 3. | code. Job Runs Over 100% faster – Now in | |
| 4. 5. 6. Key to our Approach: allowing | 1.5 hours. Data Storage: Mainframe DB2 | ||
| users to continue to use familiar | Tables. Price Data: 500M Records. | ||
| consumption interfaces providing inherent | Processing Window: 3.5 Hours. Mainframe | ||
| HA enabling businesses to unlock | Jobs: 64. | ||
| previously unusable data. | 19 | The Challenge – Use-Case #4. Hadoop. | |
| 8 | The Architecture. Enterprise solutions | Needed to enhance user experience and | |
| using Hadoop must be an eco-system Large | ability to perform analytics at granular | ||
| companies have a complex environment: | data Restricted availability of data due | ||
| Transactional system Services EDW and Data | to space constraint Needed to retain | ||
| marts Reporting tools and needs We needed | granular data Needed Excel format | ||
| to build an entire solution. | interaction on data sources of 100 | ||
| 9 | The Sears Holdings Architecture. | millions of records with agility. Teradata | |
| 10 | PiG/Hadoop Ecosystem. MetaScale. | via Business Objects. Transformation: On | |
| 11 | The Learning. Over two years of Hadoop | Teradata. User Experience: Unacceptable. | |
| experience using Hadoop for Enterprise | Batch Processing Output: .CSV Files. | ||
| legacy workload. We can dramatically | History Retained: No. New Report | ||
| reduce batch processing times for | Development: Slow. | ||
| mainframe and EDW We can retain and | 20 | The Result – Use-Case #4. Hadoop. | |
| analyze data at a much more granular | Business Problem: Needed to enhance user | ||
| level, with longer history Hadoop must be | experience and ability to perform | ||
| part of an overall solution and | analytics at granular data Restricted | ||
| eco-system. We can reliably meet our | availability of data due to space | ||
| production deliverable time-windows by | constraint Needed to retain granular data | ||
| using Hadoop We can largely eliminate the | Needed Excel format interaction on data | ||
| use of traditional ETL tools New Tools | sources of 100 millions of records with | ||
| allow improved user experience on very | agility. Sourcing Data Directly to Hadoop. | ||
| large data sets. We developed tools and | Transformation Moved to Hadoop. User | ||
| skills – The learning curve is not to be | Experience Expectations Met. Redundant | ||
| underestimated We developed experience in | Storage Eliminated. Business’s Single | ||
| moving workload from expensive, | Source of Truth. Datameer for Additional | ||
| proprietary mainframe and EDW platforms to | Analytics. PIG Scripts to Ease Code | ||
| Hadoop with spectacular results. HADOOP. | Maintenance. Granular History Retained. | ||
| IMPLEMENTATION. UNIQUE VALUE. | Over 50 Data Sources Retained in Hadoop. | ||
| 12 | Some Examples. Use-Cases at Sears | Teradata via Business Objects. | |
| Holdings. | Transformation: On Teradata. User | ||
| 13 | The Challenge – Use-Case #1. Intensive | Experience: Unacceptable. Batch Processing | |
| computational and large storage | Output: .CSV Files. History Retained: No. | ||
| requirements Needed to calculate item | New Report Development: Slow. | ||
| price elasticity based on 8 billion rows | 21 | Summary of Benefits. | |
| of sales data Could only be run quarterly | 22 | Summary. Hadoop can revolutionize | |
| and on subset of data – Needed more often | Enterprise workload and make business | ||
| Business need - React to market conditions | agile Can reduce strain on legacy | ||
| and new product launches. Sales: 8.9B Line | platforms Can reduce cost Can bring new | ||
| Items. Price Sync: Daily. Elasticity: | business opportunities Must be an | ||
| 12.6B Parameters. Offers: 1.4B SKUs. | eco-system Must be part of an data overall | ||
| Items: 11.3M SKUs. Stores: 3200 Sites. | strategy Not to be underestimated. | ||
| Timing: Weekly. Inventory: 1.8B rows. | 23 | The Horizon – What do we need next? | |
| 14 | The Result – Use-Case #1. Hadoop. | Automation tools and techniques that ease | |
| Business Problem: Intensive computational | the Enterprise integration of Hadoop | ||
| and large storage requirements Needed to | Educate traditional Enterprise IT | ||
| calculate store-item price elasticity | organizations about the possibilities and | ||
| based on 8 billion rows of sales data | reasons to deploy Hadoop Continue | ||
| Could only be run quarterly and on subset | development of a reusable framework for | ||
| of data Business missing the opportunity | legacy workload migration. | ||
| to react to changing market conditions and | 24 | www.metascale.com Follow us on Twitter | |
| new product launches. Price elasticity | @LegacyModernizationMadeEasy Join us on | ||
| calculated weekly. 100% of data set and | LinkedIn: | ||
| granularity. Meets all SLAs. New business | www.linkedin.com/company/metascale-llc. | ||
| capability enabled. Sales: 8.9B Line | Legacy Modernization Made Easy! For more | ||
| Items. Price Sync: Daily. Elasticity: | information, visit: Contact: Kate Kostan | ||
| 12.6B Parameters. Offers: 1.4B SKUs. | National Solutions | ||
| Items: 11.3M SKUs. Stores: 3200 Sites. | Kate.Kostan@MetaScale.com. | ||
| Timing: Weekly. Inventory: 1.8B rows. | 25 | ||
| 15 | The Challenge – Use-Case #2. Hadoop. | ||
| Modernizing Business with BIG DATA.ppt | |||
Modernizing Business with BIG DATA
«Data Mining» - Процесс конструирования. Простота использования полученных результатов. Продолжение. Задачи Data Mining. Области применения Data mining. Перспективы метода и выводы. Мультидисциплинарность. Возникновение Data Mining. Понятие Data Mining. Докладчики. Дескриптивный анализ и описание исходных данных. Основная идея - разделение выборки данных на v "складок".
«Windows Small Business» - © Корпорация Майкрософт (Microsoft Corporation), 2005. «Эффективное и профессиональное общение с клиентами!» Марк заволновался. Задача: работать лучше, а не больше. Лицензия «на устройство» позволяет нескольким пользователям подключаться к серверу с одного устройства. Все права защищены. Источники дополнительных сведений.
«Victory day» - Population North Korea – 23 million. Our obeisance and best wishes to the live! A shroud of a sadness still lives in our eyes. Congratulations to the veterans and their ancestors on the Great Victory Day ! Population Germany – 82million. Learn the lessons of history. Victory Day (9 May). World will not survive WW3.
«Названия дней недели» - Saturday. Это интересно знать. Wednesday. Thursday. Monday. Боги, которым поклонялись саксонские предки британцев. Понедельник. Tuesday. Sunday. Friday. Происхождение обозначений дней недели. The days of the week.
«Слогоделение английский» - Реформатский А.А. (1900-1978). Что такое транскрипция и чем отличаются английская и русская транскрипции. Проект. [Аннотация]. [Что есть слог?]. Что отличает английское слогоделение от русского? На тему: . В английском алфавите всего шесть гласных букв: a [ei],e [i], i [ai], o [ou], u [ju], y [wai].
«The green movement» - Several active workers managed to steal up on a raft to a platform and to chain themselves to it. Their features. It became the first African who has headed this organization. Green color which is used by participants of movement as the general emblem, serves as a symbol of the nature, hope and updating.
