Audience
This course is ideal for developers looking to improve their understanding of, and experience with, Apache Accumulo.
Prerequisites
Attendees must have substantial prior experience using Hadoop, and a very strong understanding of Java data access techniques.
Duration
3 days.
Course Objectives
Apache Accumulo is a Google BigTable-inspired datastore built around highly distributed key/value stores. As a part of the Hadoop ecosystem, it targets tera-scale and peta-scale datasets while maintaining redundancy, throughput, and high availability. Accumulo allows for cell-level access control as well as server-side functions for filtering and aggregation, sharing, and a pluggable balancing mechanism.
Course Content
Review of MapReduce with Hadoop
Overview of Accumulo
- Features
- History
- NoSQL
- Comparisons to other NoSQL solutions
- Hello, Accumulo!
Iterators
- Filtering
- Aggregation
Table Management
- Merging
- Cloning
- Compacting
- Deleting
Cell Labels
- Cell Labels
- Cell Access
Implementing Table Constraints
- Using Constraints
- Customizing Constraints
Sharding
- Document Sharding
- Dataset Sharding
Testing
- Unit Testing
- Mocking Accumulo
- Functional Tests
- Scaling Tests
- Random Walks
Extensibility
- Pluggable Balancers
- Pluggable Memory Manager
- Pluggable Logger Assignment Strategy