Feature/benchmark config from env #15782
Merged
+42
−23
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
ctsk
force-pushed
the
feature/benchmark-config-from-env
branch
from
571965f to
beb3a0d
Compare
|
Thanks, I think it's good to go after adding some simple docs for such configurations, in: I think it's also great if we could add some output to show the actual set configurations like Terminal output: We can just open an issue for it now, and leave it as a follow-up. |
Orchestrates running benchmarks against DataFusion checkouts Usage: ./bench.sh data [benchmark] [query] ./bench.sh run [benchmark] ./bench.sh compare <branch1> <branch2> ./bench.sh venv ********** Examples: ********** # Create the datasets for all benchmarks in /Users/christian/MA/datafusion/benchmarks/data ./bench.sh data # Run the 'tpch' benchmark on the datafusion checkout in /source/datafusion DATAFUSION_DIR=/source/datafusion ./bench.sh run tpch ********** * Commands ********** data: Generates or downloads data needed for benchmarking run: Runs the named benchmark compare: Compares results from benchmark runs venv: Creates new venv (unless already exists) and installs compare's requirements into it ********** * Benchmarks ********** all(default): Data/Run/Compare for all benchmarks tpch: TPCH inspired benchmark on Scale Factor (SF) 1 (~1GB), single parquet file per table, hash join tpch_mem: TPCH inspired benchmark on Scale Factor (SF) 1 (~1GB), query from memory tpch10: TPCH inspired benchmark on Scale Factor (SF) 10 (~10GB), single parquet file per table, hash join tpch_mem10: TPCH inspired benchmark on Scale Factor (SF) 10 (~10GB), query from memory cancellation: How long cancelling a query takes parquet: Benchmark of parquet reader's filtering speed sort: Benchmark of sorting speed sort_tpch: Benchmark of sorting speed for end-to-end sort queries on TPCH dataset clickbench_1: ClickBench queries against a single parquet file clickbench_partitioned: ClickBench queries against a partitioned (100 files) parquet clickbench_extended: ClickBench "inspired" queries against a single parquet (DataFusion specific) external_aggr: External aggregation benchmark h2o_small: h2oai benchmark with small dataset (1e7 rows) for groupby, default file format is csv h2o_medium: h2oai benchmark with medium dataset (1e8 rows) for groupby, default file format is csv h2o_big: h2oai benchmark with large dataset (1e9 rows) for groupby, default file format is csv h2o_small_join: h2oai benchmark with small dataset (1e7 rows) for join, default file format is csv h2o_medium_join: h2oai benchmark with medium dataset (1e8 rows) for join, default file format is csv h2o_big_join: h2oai benchmark with large dataset (1e9 rows) for join, default file format is csv imdb: Join Order Benchmark (JOB) using the IMDB dataset converted to parquet ********** * Supported Configuration (Environment Variables) ********** DATA_DIR directory to store datasets CARGO_COMMAND command that runs the benchmark binary DATAFUSION_DIR directory to use (default /Users/christian/MA/datafusion/benchmarks/..) RESULTS_NAME folder where the benchmark files are stored PREFER_HASH_JOIN Prefer hash join algorithm (default true) VENV_PATH Python venv to use for compare and venv commands (default ./venv, override by <your-venv>/bin/activate) DATAFUSION_* Set the given datafusion configuration
|
I can't see an easy way to check what environment variables were actually picked up. I've opted to add some logging to |
2010YOUY01
approved these changes
Apr 23, 2025
nirnayroy
pushed a commit
to nirnayroy/datafusion
that referenced
this pull request
May 2, 2025
* Read benchmark SessionConfig from env * Set target partitions from env by default fix * Set batch size from env by default * Fix batch size option for tpch ci * Log environment variable configuration * Document benchmarking env variable config * Add DATAFUSION_* env config to Error: unknown command: help Orchestrates running benchmarks against DataFusion checkouts Usage: ./bench.sh data [benchmark] [query] ./bench.sh run [benchmark] ./bench.sh compare <branch1> <branch2> ./bench.sh venv ********** Examples: ********** # Create the datasets for all benchmarks in /Users/christian/MA/datafusion/benchmarks/data ./bench.sh data # Run the 'tpch' benchmark on the datafusion checkout in /source/datafusion DATAFUSION_DIR=/source/datafusion ./bench.sh run tpch ********** * Commands ********** data: Generates or downloads data needed for benchmarking run: Runs the named benchmark compare: Compares results from benchmark runs venv: Creates new venv (unless already exists) and installs compare's requirements into it ********** * Benchmarks ********** all(default): Data/Run/Compare for all benchmarks tpch: TPCH inspired benchmark on Scale Factor (SF) 1 (~1GB), single parquet file per table, hash join tpch_mem: TPCH inspired benchmark on Scale Factor (SF) 1 (~1GB), query from memory tpch10: TPCH inspired benchmark on Scale Factor (SF) 10 (~10GB), single parquet file per table, hash join tpch_mem10: TPCH inspired benchmark on Scale Factor (SF) 10 (~10GB), query from memory cancellation: How long cancelling a query takes parquet: Benchmark of parquet reader's filtering speed sort: Benchmark of sorting speed sort_tpch: Benchmark of sorting speed for end-to-end sort queries on TPCH dataset clickbench_1: ClickBench queries against a single parquet file clickbench_partitioned: ClickBench queries against a partitioned (100 files) parquet clickbench_extended: ClickBench "inspired" queries against a single parquet (DataFusion specific) external_aggr: External aggregation benchmark h2o_small: h2oai benchmark with small dataset (1e7 rows) for groupby, default file format is csv h2o_medium: h2oai benchmark with medium dataset (1e8 rows) for groupby, default file format is csv h2o_big: h2oai benchmark with large dataset (1e9 rows) for groupby, default file format is csv h2o_small_join: h2oai benchmark with small dataset (1e7 rows) for join, default file format is csv h2o_medium_join: h2oai benchmark with medium dataset (1e8 rows) for join, default file format is csv h2o_big_join: h2oai benchmark with large dataset (1e9 rows) for join, default file format is csv imdb: Join Order Benchmark (JOB) using the IMDB dataset converted to parquet ********** * Supported Configuration (Environment Variables) ********** DATA_DIR directory to store datasets CARGO_COMMAND command that runs the benchmark binary DATAFUSION_DIR directory to use (default /Users/christian/MA/datafusion/benchmarks/..) RESULTS_NAME folder where the benchmark files are stored PREFER_HASH_JOIN Prefer hash join algorithm (default true) VENV_PATH Python venv to use for compare and venv commands (default ./venv, override by <your-venv>/bin/activate) DATAFUSION_* Set the given datafusion configuration * fmt
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Which issue does this PR close?
Rationale for this change
When benchmarking, it is convenient to modify some datafusion settings between runs and compare the results.
What changes are included in this PR?
This PR makes the benchmark executables pick up the environment variables. Additionally, it changes the target_partitions and batch_size options to fall back on the datafusion defaults when not set - this was necessary so that environment variables can override those options if they weren't provided.
Are these changes tested?
No.
Are there any user-facing changes?
Only developers of datafusion are affected.