Simple, Robust Access to Hadoop Data
Loom's API is the best point of access to Hadoop data for external tools
- Publish metadata and data through Loom API
- Provide simple access point into Hadoop for existing tools
- Out-of-the-box integration with R
For Hadoop to become a critical component of an enterprise architecture, many different existing technologies must be able to load data into HDFS and access data from HDFS. Core Hadoop provides little facility for this, beyond copying files into the file system. Hive provides another point of access, but it too is limited.
Loom provides a centralized point in Hadoop for metadata management and data access. All files and directories are registered automatically by Activescan as Datasets. Activescan generates schemas and statistics for all Datasets. Loom's registry is extensible, so any other metadata about datasets can also be registered. Loom provides mechanisms for sampling and transforming datasets.
Loom's RESTful API offers a single point for any third-party application to access metadata about any Hadoop data. The API also enables third-party applications to directly access data or execute transformations in the cluster.