Cade Mirchandani
Cade Mirchandani
A lot to look at here. I believe there is an "inventory" of files built before the DAG is constructed, but it seems that may not be efficiently built for...
I've taken a look at the other two issues you've opened. Since the core issue isn't specific to one storage plugin, let's consolidate the discussion here. To answer your questions:...
> Thanks for consolidating this here and for pointing to the specific code in `job_scheduler.py`. That `update_input_sizes` method using `asyncio.gather` with individual `path.exists()` and `path.size()` calls is exactly where I'm...
This should be fixed by #3881.
Will review asap - this looks great at first glance. Thank you @jlumpe. A few thoughts: > #### Global logger > There is still a single global `Logger` instance, which...
If you use`--cores 1` does that run 1 job at a time? While the docs say jobs is an alias for cores when executing locally, it does suggest using cores....
I'm a bit confused - when you run a workflow with a rule that has a `conda:` directive and you don't supply `--use-conda`, Snakemake creates the environment anyways? If that...
Hmm. Do you have any thoughts on how to add it back in?