Back to Volume
Paper: Data Versioning in the Chandra Data Archive
Volume: 351, Astronomical Data Analysis Software and Systems XV
Page: 711
Authors: Zografou, P.; Harbo, P.; Patz, A.; Tibbetts, M.; Van Stone, D.
Abstract: The Chandra Data Archive stores multiple versions of data products generated during processing and reprocessing of observation data. Versions of a product are variants of a processing pipeline output generated in different processing runs. A version is identified by a number which is determined by the pipeline. It is related to other versions of the same product by a set of criteria. All versions of Chandra data products are stored in the archive and are available for retrieval. In this paper, we describe how a version number is generated and how product versions in the archive are identified for the various data types. We present how multiple versions of products are stored and how they are accessed by users. Finally we follow the lifetime of a version in the archive from ingestion to statuses pending, default and demoted with the resulting changes in location and access by search and retrieval.
Back to Volume