Kevin German

It would be helpful if the Conda Environment plugin were able to update the results#requirements with the exact conda version of a dependency rather than the conda refspec that was...

file names should not use externally generated strings

1

Currently result files are named using the hash that was benchmarked. For repos which use an actual hash as an identifier this has been working, however, if the repo impl...

recursive serialization fails with nested objects

11

When serializing nested more than 511 dictionaries deep, `msgpack`, used by `distributed.protocol.core.dumps` hits a hard coded recurrsion limit. ``` distributed.protocol.core - CRITICAL - Failed to Serialize Traceback (most recent call...

When running multiple instances of GPU_BDB from the same working directory the interim results files, `benchmarked_*.csv` will be overwritten and/or corrupted. Add the run_id configuration option into each filename to...

WIP: log dask task stream and rmm events into results dir

1

Adds a couple of utilities to capture metrics per query triggered in benchmark_runner.py rmm_monitor logs Cuda allocations from each dask worker dask_task_logger grabs the tasks into a json file TBD:...