Speaker
Description
Over the past several years, the dCache collaboration has been working on developing a feature-rich, efficient and scalable data lifecycle and QoS service. With the ever-increasing volume of data anticipated by the LHC and Intensity Frontier experiments, their reliance on tape storage and the limited amount of disk cache available to them, this effort to provide for efficient staging of large data samples from tape to disk and management of data lifecycles has become even more timely and urgent.
The major dCache release of 2020 (6.2) included a prototype of the Bulk service, which supports the concurrent processing of multiple file targets as well as the recursive handling of directory targets for such activities as pinning or restoring files to disk, deleting files and changing their quality of service requirements. With the major release of 2022 (8.2), that prototype has been significantly improved for scalability, reliability and efficiency. Concomitantly with these changes, dCache's native RESTful interface, which provides the front-end user access to the Bulk service, has also been expanded to handle the specific REST resources required by the WLCG tape API specification (https://docs.google.com/document/d/1Zx_H5dRkQRfju3xIYZ2WgjKoOvmLtsafP2pKGpHqcfY/edit#heading=h.ozszs1lr7q93).
This presentation will describe the dCache Bulk service and its RESTful façade, and will demonstrate its capabilities.
| Consider for long presentation | No |
|---|
