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
We have multiple (nearly) static collections that are being regenerated which are actually quite large, which is taking quite long and creating many items.
It would be great if we could make use of this, i think there are multiple potential approaches.
Potentially the most straight forward one is for large time series collections instead of creating items individual items, we create one geoparquet. The caveats i foresee is:
the bubbled up information would not be available so we would need a dedicated handler in eodash to fetch important information from geoparquet (not sure if it is possible to extract only one property?)
how would we stop pytac from actually trying to save results when saving the catalog?
maybe we can "manually" create the geoparquet and add the links reference and don't actually add the items to the collection
Other integration would be to somehow be able to set if geoparquet should be used as output format, potential caveats:
should we consider an update logic approach? e.g. if file is already present only update if update flag?
(same issues in the client as in 1)
Any other considerations? Let's brainstorm!
Inputs:
We have multiple (nearly) static collections that are being regenerated which are actually quite large, which is taking quite long and creating many items.
It would be great if we could make use of this, i think there are multiple potential approaches.
Any other considerations? Let's brainstorm!
Inputs:
The text was updated successfully, but these errors were encountered: