You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
There are twin issues here. One is that for the DataSM, any mechanical failure in "visual validation" would unnecessarily cause a break in the entire "Postprocess" (derivative datasets) workflow. As these visuals were a subordinate stage of Postprocess, the only (DataSM) way to recover would be to re-run the Postprocess jobs (re-extract and cmorize). The alternative (manual intervention) requires editing each status file to add "Postprocess:Pass" to cover the preceding "Fail".
What we need are two things:
A suite of automated "data validation" measure that can be applied to flag real data problems automatically.
A suite of automated visual generations as an independent, top-level Pre-publication stage.
Both the data-based and visual-based validations can be invoked as part of the DataSM operations. Perhaps neither should be in the critical path of Postprocessing, but an independent follow-on stage.
This workflow needs to be brought back by plotting tools updates. Right now manually point check is relied on before publication.
The text was updated successfully, but these errors were encountered: