-
Notifications
You must be signed in to change notification settings - Fork 5
Home
Unlike all previous benchmarks which are either micro benchmarks or benchmarks for very specific domains, BigFrame is actually a Benchmark Generator targeting next-generation big data analytics. Given BigFrame, a user now can generate her own benchmark tailored to her special needs, according to the Data variety, volume and velocity she specified.
We thank the US National Science Foundation for supporting this project!
There is a small corpus of helpful documents on BigFrame. The bigframe-users mailing list is a helpful resource for beginners
- Slides for BigFrame (A brief introduction about BigFrame is presented in Hadoop Submit 2013.)
- Paper describes our main idea
- BigFrame User Guide A tutorial talks about how to use BigFrame.
BigFrame provides a variety of workflows involving different combinations of data types, which are characterized by specific application domains (in short, app domains). The current version has implemented a Business Intelligence app domain, which provides relational data, nested textual data, and graph data. We will implement more app domains in the future. Users can pick the workflows to benchmark their systems against as per their need. Listing below some of the possible workflows categorized based on data they process. This, by no means, is an exhaustive list and we will keep adding more workflows.
- Relational data only
- Text data only
- Mix of relational and text data
- Mix of relational, text and graph data
For people who are interested in contributing: