The Globus Search service enables storage of metadata with fine grained access control on creation, modification, and visibility, and metadata retrieval through search queries. Globus Search is a set of capabilities that can be incorporated into your data management applications, either in conjunction with, or independently of, other Globus features such as file transfer and data sharing. The starting point is to provide Globus Search with metadata. Unlike many other services, there is no prescribed "metadata schema"; Globus Search will ingest metadata of arbitrary structure and make it available for faceted search. The primary element of the service is the index, which is the vehicle for storing metadata, setting permissions, and performing searches.
Globus Search metadata are organized using "subjects" and "entries". Entries allow association of multiple distinct pieces of metadata with a single subject, making it possible to have, for example, multiple sources contributing metadata independently of each other or different visibility policies on different entries (such as may be required in an environment that includes both open and protected data). Queries in Globus Search utilize ElasticSearch and allow for both basic text matching (e.g. a search field in a web application) and more advanced syntax that incorporates ranges, regular expressions, matching on particular fields, and other more sophisticated capabilities (e.g. to facilitate more targeted discovery in data portals and science gateways).
Explore Globus Search
- Check out the searchable demo application that shows how Globus Search can be used to build an index of file metadata
- The Search CLI is a command-line tool for interacting with Globus Search
- Globus SDK for Python: Documentation · Source Code
- Experiment with the platform via Jupyter notebooks
- Globus Search API Reference