Allow for scheduler file and existing dask cluster when using pdsh#21024
Allow for scheduler file and existing dask cluster when using pdsh#21024rapids-bot[bot] merged 15 commits intorapidsai:mainfrom
Conversation
python/cudf_polars/cudf_polars/experimental/benchmarks/utils.py
Outdated
Show resolved
Hide resolved
python/cudf_polars/cudf_polars/experimental/benchmarks/utils.py
Outdated
Show resolved
Hide resolved
python/cudf_polars/cudf_polars/experimental/benchmarks/utils.py
Outdated
Show resolved
Hide resolved
Co-authored-by: Lawrence Mitchell <wence@gmx.li>
rjzamora
left a comment
There was a problem hiding this comment.
I think this change makes sense.
In follow-up work, I'd like to move the more worker-setup logic (e.g. rmm resource creation) into rapidsmpf. For example, something like rapidsai/rapidsmpf#779 will allow us to setup the rmm mr at bootstrapping time.
python/cudf_polars/cudf_polars/experimental/benchmarks/utils.py
Outdated
Show resolved
Hide resolved
wence-
left a comment
There was a problem hiding this comment.
Two small cleanup suggestions
python/cudf_polars/cudf_polars/experimental/benchmarks/utils.py
Outdated
Show resolved
Hide resolved
python/cudf_polars/cudf_polars/experimental/benchmarks/utils.py
Outdated
Show resolved
Hide resolved
Co-authored-by: Lawrence Mitchell <wence@gmx.li>
Co-authored-by: Lawrence Mitchell <wence@gmx.li>
|
/ok to test 7fa5a2c |
|
/merge |
|
/ok to test 9c83470 |
| if scheduler_address is not None: | ||
| # Connect to existing cluster via scheduler address | ||
| client = Client(address=scheduler_address) | ||
| n_workers = len(client.scheduler_info().get("workers", {})) |
There was a problem hiding this comment.
Just a note: we serialize run_config.n_workers in the JSON output. When a scheduler file is provided the run_config.n_workers won't be accurate.
I wonder if we can mutate run_config.n_workers here? It's not ideal, but I think it's an OK tradeoff.
|
/ok to test f422439 |
|
Merged with main which should hopefully resolve [doc errors were seeing](https://github.com/rapidsai/cudf/actions/runs/21073573023/job/60614010671?pr=21024#step:13:5132 |
|
/ok to test 7481a1e |
|
I think this is failing now because 26.4 packages aren't built yet: |
|
/ok to test |
@pentschev, there was an error processing your request: See the following link for more information: https://docs.gha-runners.nvidia.com/cpr/e/1/ |
|
/ok to test fb91090 |
cc @rjzamora