VoltDB 소개
1 like453 views
The document outlines VoltDB, a high-performance in-memory relational database designed for fast data processing and real-time analytics, emphasizing its scalability, ACID transaction support, and compatibility with SQL and Java. It highlights major clients like LG, Huawei, and Samsung, showcasing its capabilities in handling millions of streaming data transactions. The architecture supports various applications and data sources, achieving low-latency responses and efficient operational analytics.
VoltDB 소개
- 1. 1 16-Jun-18 Head of Sales @ VOLTDB jlee@VOLTDB.com +82 10 9196 1416 In-Memory based RDBMS / 100X Faster / ACID Transaction / Built-in HA / SQL+Java support In-Memory (FAST DATA) Event
- 2. 2 “ In-Memory 기반 고성능 데이터베이스” “초당 수백만건의 Streaming 데이터 처리 및 실시간 분석” “ LG, 화웨이, 노키아, 미쯔비시 등 고객사 다수 “ “ 삼성전자를 포함해서 국내 고객사( )등에서 성능/기능 검증 됨” ▪ In-Memory ▪ Scale-Out, Shared Nothing Cluster ▪ Streaming ▪ Global - , LG, Huawei, Intel, Nokia, Mitsubishi ▪ IT / – Stock/IoT/Game/E-commerce https://www.voltdb.com/
- 3. 3 https://www.VOLTDB.com/resources/videos-technical/ FAST DATA ? FAST DATA , BIG DATA , Transaction , , , BIG DATA
- 4. 4 https://www.VOLTDB.com/resources/videos-technical/ FAST DATA ? FAST DATA Sensor , Transaction , , Event Event Sensors Click Streams Machine Data Data from Device Events Transactions Customer Interactions (Ingestion) (Analytics) Export To BigData FAST DATA Processing
- 5. 5 VOLTDB In-MemoryIn-MemoryIn-Memory Export ApplicationsMessageQueuesDataSources … Data Warehouse ▪ 100% In-Memory ▪ Scale-Out Cluster ▪ Shared Nothing Distributed ▪ ACID Relational Database SQL ▪ Event OLAP/Big Data ▪ Fast Data ▪ (Authorization) & (Authentication) ▪ (Personalization) ▪ / (Fraud Detection/Correction) ▪ Event VOLTDB VOLTDB Intelligence Feedback
- 6. 6 VOLTDB FAST DATA ACID SQL Relational DB(RDB) Processing Relational State(RDB)VOLTDB VOLTDB ACID In-Memory New SQL (RDB ) , FAST DATA
- 7. 7 VOLTDB - VOLTDB FAST DATA DB (NEW SQL)
- 8. 8 Kafka / RabbitMQ Storm, Flume, Sqoop Storm + Serving Layer Spark + Serving Layer Cassandra, HBase Hadoop, Message queues VOLTDB - Open Source vs. VOLTDB Applications, Message Queues, Data Sources Ingest Analyze Decide • Counters • Aggregations • Time Series • Statistics • Store Results • Query/Recombine • Fast Serving • Per-event policy evaluations • Responses (synchronous) • Side-effects (asynchronous) Export & Pipeline Open-source - Open-source -- FAST DATA / Generic -
- 9. 9 Kafka / RabbitMQ Ingest Interface SQL, Java for Analytics Transactions / ACID Hadoop, Message queues - VOLTDB -- FAST DATA / Generic - VOLTDB VOLTDB - Open Source vs. VOLTDB Applications, Message Queues, Data Sources Ingest Analyze Decide • Counters • Aggregations • Time Series • Statistics • Store Results • Query/Recombine • Fast Serving • Per-event policy evaluations • Responses (synchronous) • Side-effects (asynchronous) Export & Pipeline
- 10. 10 Kafka / RabbitMQ Ingest Interface SQL, Java for Analytics Transactions / ACID Hadoop, Message queues Kafka / RabbitMQ Storm, Flume, Sqoop Storm + Serving Layer Spark + Serving Layer Cassandra, HBase Hadoop, Message queues - VOLTDB -- Open-source - Component Transaction vs. VOLTDB - Open Source vs. VOLTDB Opensource Component Interop
- 11. 11 FAST DATA VOLTDB VOLTDB FAST DATA / Big Data Ingest / Interactive Analytics Decisioning Fast Operational Database Enterprise Apps ETL CRM ERP Etc. Data Lake (HDFS) BIG DATA Non Relational Processing BI Reporting Data Warehouse Columnar Analytics OLAP Export FAST DATA DW EXPORT DW Materialized View
- 12. 12 VOLTDB - Real-time Analytics Transactions Transformations • Materialized Views • Capped Tables • Ranking Indexes • Per-event Java + SQL • ACID processing • Millisecond latency responses. • Loaders/Importers • Export Connectors • Sessionization • Enrichment VOLTDB Architecture Commodity HW HA + ACID Scale-out VM-friendly VOLTDB FAST DATA /
- 13. 13 2014 ACM Turing Award Mike Stonebraker Stonebraker : / • In-Memory (durable to disk) • Scale-out Shared-nothing • • Hadoop Data warehouse • Open source Cloud : YCSB (Yahoo Cloud Serving Benchmark) 2.4m TPS HERITAGE INNOVATION SUCCESS
- 14. 14 ▪ Michael Stonebraker - RDBMS , 1973 Ingres & PostgreSQL › UC Berkeley - RDMS B-trees, Constraints, Trigger , Sybase, Microsoft SQL Server - 2006 : “ DB RDBMS Architect ” - DB in-memory New SQL (VOLTDB) - , MIT - Michael Stonebraker(Founder of VOLTDB)
- 15. 15 - One transaction = One Client-DB interaction - Java Stored Procedures Multi Statement Transaction High-Velocity In-Memory ACID SQL In-Line - Import / (analytics/decisions) - Transaction - Disk Durability - RDBMS , 100x H/W - Scale Out - - - - SQL - Cluster - Built-in Failover Replication - Cloud - Shared Nothing VOLTDB
- 16. 16