⚠️ CI failed ⚠️ - regressions: dataframe_cow_chain - prepreocess - q6 - q8 - set_index, write_wide_data

[github-actions[bot]]

Workflow Run URL

[milesgranger]

Most don't seem particularly worrisome.

	category	type	mean	last	last-1	last-2	threshold
('coiled-upstream-py3.9', 'test_dataframe_cow_chain', 'duration') [s]	benchmarks	duration	31.0223	39.4496	39.7347	39.0878	32.0223
('coiled-upstream-py3.9', 'test_dataframe_cow_chain', 'average_memory') [GiB]	benchmarks	average_memory	17.8715	18.9385	19.0462	18.9521	18.8715
('coiled-upstream-py3.9', 'test_preprocess', 'peak_memory') [GiB]	benchmarks	peak_memory	28.2098	35.8898	35.766	35.4389	30.5931
('coiled-upstream-py3.9', 'test_q6[5 GB (parquet)-tasks]', 'peak_memory') [GiB]	benchmarks	peak_memory	11.1026	12.2275	12.2244	12.2719	12.1974
('coiled-upstream-py3.9', 'test_q8[5 GB (parquet)-tasks]', 'peak_memory') [GiB]	benchmarks	peak_memory	12.7532	14.0084	13.9291	13.9719	13.7532
('coiled-upstream-py3.9', 'test_set_index[0.1-False]', 'peak_memory') [GiB]	benchmarks	peak_memory	11.1966	12.7245	12.6605	12.9963	12.3368
('coiled-upstream-py3.9', 'test_write_wide_data', 'duration') [s]	benchmarks	duration	185.905	204.313	203.649	209.776	195.489
('coiled-upstream-py3.9', 'test_write_wide_data', 'average_memory') [GiB]	benchmarks	average_memory	46.3999	53.3078	53.8739	53.0613	47.6307
('coiled-upstream-py3.9', 'test_write_wide_data', 'peak_memory') [GiB]	benchmarks	peak_memory	64.7377	76.3773	76.4972	76.5402	70.9022

cc @phofl can we confidently ignore these you think?

[fjetter]

@milesgranger can you look up in the benchmark database what version change triggered this?

[milesgranger]

Looks like the two versions that changed at the point of regression were distributed==2024.3.0+4.g07489d0d and dask-expr==1.0.2

db-view-test_write_wide_data-duration-regression.csv