Our team has zeroed-in to Couch DB as their choice for NoSQL persistence storage.
Below is the gist of session where we gave quick glance of the important Couch DB features that matter the most for Game Integrity team.
Summary of Session
1. Problems with traditional RDBMS Vs. No SQL Promises
- Explicit ORM Mapping
- Schema Upgrades and maintenance
- Downtime for schema upgrade
- PL/SQL portability issue
- Query / Views flexibility (Index Vs Map-Reduce)
- Ease of Use (HTTP / JSON )
2. Technical Overview
- Append only data, Immutability of data, leveraging Functional Programming platform (Erlang)
- MVCC (Multi Version Concurrency Control) using versioning
- Eventual Consistency, Conflict handling
- Data Replication using master to master strategy
3. Introduction to Map Reduce (Design / Views)
- Map (Data Filtering)
- Reduce (Data Aggregation and Statistics)
- Map-Reduce Sample using java-script
4. Bots Couch DB Usage and its production deployment strategy
- Nature of Bots Data and performance requirements
- Clubbing Jar Deployment along with NoSQL Artefacts
- View Upgrade / Deployment (Index Upgrades)
Rahul Dongare 