1 / 12

HBase: Under The Hood

HBase: Under The Hood. Asaf Mesika Senior Engineer / Akamai. Goals. Columnar DBs Motivation. How does HBase work?. HBase Use Cases. Agenda. RDBMS to NoSQL – Why? (motivation) HDFS 101 HBase 101 Schema Architecture Production war stories (If time permits)

grant
Download Presentation

HBase: Under The Hood

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript


  1. HBase: Under The Hood Asaf Mesika Senior Engineer / Akamai

  2. Goals Columnar DBs Motivation How does HBase work? HBase Use Cases

  3. Agenda • RDBMS to NoSQL – Why? (motivation) • HDFS 101 • HBase 101 • Schema • Architecture • Production war stories (If time permits) • Akamai Security Analytics in numbers • Problems and matching home brew solutions

  4. Projects got bigger – Web Facing • My Web is a success – lifecycle • HBase is the solution RDBMs -> NoSQLWhy?

  5. Why HBase is the solution? • Scale Horizontally • Common Hardware (cheap) • Strongly consistent read/write • Automatic Sharding and LB • Automatic Failover

  6. HDFS • Traditional File System • Distributed File Systems • Motivation • File Structure in HDFS • HDFS Goal: Streaming

  7. HDFS Architecture • Components: NameNode, DataNode, Secondary NameNode, Balancer • NameNode • DataNode • Client Flow (+Replication Pipelining) • Replication Placement Policy (+Rack Awareness) • Robustness

  8. HBase – Table Schema • RDBMS Table • HBase Table • Transitioning to Key-Value way of thinking

  9. HBase Architecture • Table = Set of Regions • Components – Master, Region Server, Zookeeper • Region Server: zoom-in • Master: zoom-in

  10. Production use-case: Akamai • 30 RS, each with 8 x 2TB HD, 32 cores, 32GB RAM, 1 Gbit Ethernet • Increments Throughput – JAFI • No Query Language - Yoda

  11. HBase vs. other solutions • Cassandra • Eventual Consistency • Same problematic support of atomic increments • Hashed (evenly) distribution • Document oriented • MongoDB, CouchDB

  12. I want it, what’s next? • Book • HBase The Definitive Guide • Start at least with 4 computers • Installation Quick-start guide available at http://devving.com • Browse/Subscribe to the official mailing list • They discussed almost any design/installation/bugs issues out there

More Related