Defining and exercising the full set of steps required to go from raw data to end user products and their service to users is important for understanding where the gaps are in the operational and software systems and how to estimate key performance metrics.
End-to-end testing in 2017 will extend to include both additional Science Platform functionality and a new processing of precursor data (the first HSC public data release) with the LSST stack.
dbserv and imgserv APIs, and taking advantage of metaserv once it becomes available. Please see LSP Integration and Prototype Deployment for more details.End-to-end testing in Fall 2016 and into early 2017 is based around the integration of database services (Qserv and others), DAX services above them (dbserv, imgserv, and later metaserv), an initial version of the SUIT Portal components (services and pages), and the NCSA-provided integration-cluster infrastructure. The data used for this testing are previously-processed datasets (the LSST D&D-project's Summer 2013 processing of SDSS Stripe 82, and the IPAC processing of the WISE mission data) - this round of testing does not include current LSST-stack data processing.
Initial prototyping will be done at the Bremerton meeting in August. The aim is to do as complete an end to end test as we can manage although it does not have to be completely automated; stages can have manual intervention. The aim it to be able to determine the areas where we are most lacking in terms of a full end-to-end test (a full test from simulated OCS/Camera to Alert Production is defined in ) and to make initial estimates for our Key Performance Metrics.
