1 / 9

FaceBase Data Management

FaceBase Data Management. Data Management Goals. Schema-less data format Fast search Fast update Scalability Transactions* Atomicity Consistency Isolation Durability. Having it all…. No one system has everything Brewer’s theorem (CAP) Consistency Availability Partition Tolerance

chana
Télécharger la présentation

FaceBase Data Management

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. FaceBase Data Management

  2. Data Management Goals • Schema-less data format • Fast search • Fast update • Scalability • Transactions* • Atomicity • Consistency • Isolation • Durability

  3. Having it all… • No one system has everything • Brewer’s theorem (CAP) • Consistency • Availability • Partition Tolerance • What can we compromise on? • What can we sacrifice?

  4. MongoDB • Document store • BSON • Search: Fast, but indexes have some restrictions • Prefixing – no true multi-dimensional indexes • Update: Fast, but indexes have to explicitly declared • Scalabilty • Designed to scale out and shard • Strong consistency and eventual consistency

  5. Solr • Document store • Each document has several key-values, no hierarchy • Web API • Search • Fast keyword and range searches • Mulit-attribute search? • Update • No updates searchable until “commit” • “commit” is time-consuming – rebuild the index • Scalability • Similar to “commit”, replication requires creating a complete image

  6. Cassandra • Document store • Indexed on primary key – key-value model • Fast single-key equality search • More complex queries not as efficient • Fast updates • Scalability • Designed for distributed environment • Emphasis of high availability, partition tolerance • Brewer’s Conjecture: consistency is compromised

  7. MySQL • Relational model • Build a document model over relational model • Fast search • Fast update • Performance concerns with large data set • Scalability • Replication, sharding • True transactions • Easy to generate sequence numbers • Transactions can incorporate Drupal and FB data w/o coupling

  8. Head to Head

  9. Possible Routes • Hybrid MongoDB and MySQL • Benefits: Less resource demand • Drawback: Recovery and durability • Keep metadata in MongoDB, use MySQL as a catalogue • In-house EAV with MySQL • Benefits: Better interoperability • Drawback: Higher resource demand • Implement EAV store over MySQL.

More Related