Dyna Robotics Turbocharges Foundation Model Training with Alluxio
Interactive Analytics with Trino and Alluxio
Drive faster insights with high performance SQL analytics
.png)
Trino is an open-source, distributed SQL query engine that enables fast and interactive queries on large datasets. Alluxio enhances Trino's capabilities by providing a high-performance data access layer, ensuring application portability, unified data access, and significant cost reductions for interactive analytic workloads.
Alluxio provides seamless data access through flexible APIs and a unified namespace. Regardless of where your data resides, Alluxio makes it effortlessly accessible to Trino, while transparently caching frequently accessed data to optimize performance and speed up queries.
Alluxio + Trino = Better Together
Cross-Region Data Access
Alluxio provides data access to Trino workloads across datacenters, regions, and clouds regardless of the data source
Enhance Query Performance & Reduce I/O
Alluxio improves query speed by colocating its workers with Trino, reducing I/O latency in remote or congested networks.
Share Cache Between Workers
Alluxio serves as a shared cache between Trino workers, and across Trino clusters and other compute engines
Enable Resilient Caching
Alluxio enables resilient caching to save costs using spot instances without losing cached data
Application Portability
Alluxio makes Trino applications portable across hybrid or multi-cloud
Simplify Data Management
Alluxio provides flexible APIs and a unified namespace to simplify data management
Which Cache Fits Your Needs?
Alluxio Distributed Cache
- Need to share data across different compute frameworks
- Data spans cross-region or hybrid cloud
- Data size larger than XYZ
Alluxio Edge
- Your version of Trino is 438 or older
- Need enterprise level support
- Data size larger than XYZ
Trino Filesystem Cache
- Have a single Trino cluster
- Data is less than XYZ
- No data sharing needed between compute engines
.jpeg)
