Ryan Abernathey
Ryan Abernathey
To further clarify, listing a large bucket like this with gcsfs takes > 10 minutes. That's unacceptably long for our users. Also, we might want to deliberately control the data...
Let me suggest a path forward here. We need some kind of intake driver that tells us how to interpret the csv file. For example, this might look like ```yaml...
> Would it be useful, then, to have the `Entries` type used within a catalog to support dataframes directly? The assumed model is dictionary-like, but so long as `list(entries)` and...
xref https://github.com/pangeo-forge/pangeo-forge-recipes/issues/242#issuecomment-975891812
I recently discovered this issue (xref https://github.com/pangeo-forge/pangeo-forge-orchestrator/pull/28). In that project, we are trying to include API documentation from a fastapi api object directly in a sphinx website. I imagine this...
> For better or worse, the test suite passes with those lines removed... Drat. I know there are some examples that break without it, but I guess not the test...
I agree that we should pursue an upstream fix. But there are also a few other possibilities ### Make time coordinate contiguous from the beginning _What if we explicitly specify...
If the locks are fine-grained enough (e.g. _only_ on the `time` variable) then it should not be too bad. Async within pangeo-forge would help a lot here too--we could write...
> I'll look into doing the coordinate consolidation earlier (in `expand_target_dim`). You'll need to revisit this line: https://github.com/pangeo-forge/pangeo-forge-recipes/blob/78a274b4cceb4bc1734cddc48fa9c58f553f7684/pangeo_forge_recipes/recipes/xarray_zarr.py#L499 We currently are assuming that all variables with `time` (more generally called...
Perhaps there could also be a global config option to set `batch_size`?