perf: change vectorized equal to in multi column aggregation to use BooleanBufferBuilder instead of &[bool] [WIP]#19103
perf: change vectorized equal to in multi column aggregation to use BooleanBufferBuilder instead of &[bool] [WIP]#19103rluvaton wants to merge 14 commits intoapache:mainfrom
BooleanBufferBuilder instead of &[bool] [WIP]#19103Conversation
|
run benchmark aggregate_vectorized |
|
🤖 Hi @rluvaton, thanks for the request (#19103 (comment)).
Please choose one or more of these with |
|
@alamb anyway we can add this as well? |
|
run benchmarks |
|
🤖 |
Could you provide a PR to https://github.com/alamb/datafusion-benchmarking ? |
|
🤖: Benchmark completed Details
|
|
run benchmarks |
|
🤖 |
|
🤖: Benchmark completed Details
|
|
run benchmarks |
|
🤖 |
|
run benchmarks aggregate_query_sql |
|
🤖 Hi @rluvaton, thanks for the request (#19103 (comment)).
Please choose one or more of these with |
|
🤖: Benchmark completed Details
|
|
I think these benchmarks are grouping on bytes/bytes view columns and I need to further optimize there, but the |
|
run benchmark aggregate_vectorized aggregate_query_sql |
|
🤖 |
|
🤖 |
|
🤖: Benchmark completed Details
|
|
Show benchmark queue |
|
🤖 Hi @rluvaton, you asked to view the benchmark queue (#19103 (comment)). No pending jobs in |
|
run benchmark aggregate_vectorized |
|
🤖 |
|
run benchmark aggregate_vectorized |
|
Show benchmark queue |
|
🤖 Hi @rluvaton, you asked to view the benchmark queue (#19103 (comment)).
|
|
Show benchmark queue |
|
🤖 Hi @rluvaton, you asked to view the benchmark queue (#19103 (comment)).
|
|
@alamb I think the benchmarks are stuck or something, its been 5 hours |
|
Show benchmark queue |
|
🤖 Hi @rluvaton, you asked to view the benchmark queue (#19103 (comment)).
|
Looking |
|
The script died with this error: (aka the script is not robust to errors). I just merged this PR up from main to try and get your fix for that and restarted |
|
🤖 |
|
(actually I have an idea for a workaround...) |
|
🤖 |
|
Hoping this fixes it: alamb/datafusion-benchmarking@a3e755a |
|
🤖: Benchmark completed Details
|
|
🤖 |
|
🤖: Benchmark completed Details
|
TODO
Which issue does this PR close?
Rationale for this change
What changes are included in this PR?
Are these changes tested?
Are there any user-facing changes?