Keith has run timing tests on loading metrics into memory via the RSP
Question from first-look analysis discussions: we put information in butler, but then how do we get this information to the observer? Is there an all-done signal returned from OCPS? Do we envision routine monitoring / querying?
Commissioning test planning, based on DM experience
write tests "just-in-time".
mapping of system-level test cases to test plans and tying these to milestones
have test plans for data taking on the summit
Erik: LATISS vs ComCam dataset. For illuminated images, ComCam is a bit easier to use since AuxTel requires active summit support for clearance on slewing the telescope and turning on the lamp. ComCam is always available unless otherwise under testing/maintenance and the lamp is left on. For darks/biases, both ComCam and LATISS are regularly available.
Timing tests
30 mins to just load metrics into memory in the RSP.
With w_2021_33 on RSP, takes 10 seconds to queryDatasets for 48K measurements.
Takes 1m52s to load 10K Measurements into memory.
What about using parquet files? We don't have to use lsst.verify as the im-memory storage. Could use astropy tables. Simon thinks yaml parsing is the bottleneck. Metrics computed on small scales will inevitable result in irreducible I/O.
Discussion to date has been informal. This is really a request with a commissioning hat on.
sandbox dashboard
manage with tags in the interim
sasquatch product from Angelo will unify squash/efd.