OODT (Object Oriented Data Technology) is an open-source, data management framework currently distributed by Apache. Originally from NASA’s Jet Propulsion Laboratory, this component-oriented software was developed to focus on access to science data repositories and data generation and capture.
We sat down with Arpan Bhattacharya, Big Data Engineer here at Knowledgent, to pick his brain on the advantages of using OODT:
How did you first become interested in OODT?
We’re always looking for technologies that we can use with big data. With Hadoop specifically, we’d been thinking about integrating metadata management.
What are some of the advantages of using OODT with Hadoop for data warehousing?
OODT really is a good fit for Hadoop with regard to metadata. Hadoop does not support metadata management natively, except with the limited support in HCatalog, so by integrating OODT with Hadoop, we enable metadata management with distributed computing capabilities.
For example, when used as the entry point for a data warehousing solution, OODT can automatically extract metadata to all files ingested.
Also, when you create data sets (what OODT calls product types), you can establish default values for the metadata. This actually allows you to customize how you want to group and store the data (by project or vendor, for instance) regardless of file type. It also helps you find all the data associated with a specific data set or vendor via search. Instead of having separate repositories to store different types of files, everything is stored in one place with uniform metadata.
Is there anything else you like about OODT?
In OODT, every file that is ingested needs to be part of a data set. The default metadata can be configured using XML files, but we have used an ontology to allow a more user-friendly way of creating default metadata for data sets. Instead of using XML files, you can use a GUI to edit ontologies, which control the data set’s configuration. This enables greater flexibility and customization of metadata.
Are you using OODT? What do you think of it? Let us know in the comments!