Try fix some stuff?

Signed-off-by: Adam Gutglick <[email protected]>

Author: AdamGS

.github/workflows/sql-benchmarks.yml

@@ -73,15 +73,17 @@ on:
               "id": "fineweb",
               "subcommand": "fineweb",
               "name": "FineWeb NVMe",
-              "targets": "datafusion:parquet,datafusion:vortex,datafusion:vortex-compact,duckdb:parquet,duckdb:vortex,duckdb:vortex-compact"
+              "targets": "datafusion:parquet,datafusion:vortex,datafusion:vortex-compact,duckdb:parquet,duckdb:vortex,duckdb:vortex-compact",
+              "scale_factor": "100"
             },
             {
               "id": "fineweb-s3",
               "subcommand": "fineweb",
               "name": "FineWeb S3",
               "local_dir": "vortex-bench/data/fineweb",
               "remote_storage": "s3://vortex-bench-dev-eu/${{github.ref_name}}/${{github.run_id}}/fineweb/",
-              "targets": "datafusion:parquet,datafusion:vortex,datafusion:vortex-compact,duckdb:parquet,duckdb:vortex,duckdb:vortex-compact"
+              "targets": "datafusion:parquet,datafusion:vortex,datafusion:vortex-compact,duckdb:parquet,duckdb:vortex,duckdb:vortex-compact",
+              ""scale_factor": "100"
             },
           ]
 
@@ -260,7 +262,6 @@ jobs:
           # Run df-bench with remote storage
           if [ -n "$df_formats" ]; then
             target/release_debug/df-bench ${{ matrix.subcommand }} \
-              --use-remote-data-dir ${{ matrix.remote_storage }} \
               -d gh-json \
               --formats "$df_formats" \
               $opts \

benchmarks/df-bench/src/main.rs

@@ -75,9 +75,6 @@ struct Args {
     #[arg(long, default_value_t = false)]
     hide_progress_bar: bool,
 
-    #[arg(long)]
-    use_remote_data_dir: Option<String>,
-
     #[arg(long, default_value_t = false)]
     emit_plan: bool,
 
@@ -121,7 +118,7 @@ async fn main() -> anyhow::Result<()> {
             .to_file_path()
             .map_err(|_| anyhow::anyhow!("Invalid file URL: {}", benchmark.data_url()))?;
 
-        for format in args.formats.iter().copied() {
+        for format in args.formats.iter() {
             match format {
                 Format::OnDiskVortex => {
                     convert_parquet_to_vortex(&base_path, CompactionStrategy::Default).await?;

vortex-bench/src/runner.rs

@@ -332,6 +332,7 @@ impl SqlBenchmarkRunner {
 
             for (query_idx, query) in queries.iter() {
                 let query_idx = *query_idx;
+                tracing::debug!(%format, query_idx, "Running query");
                 self.run_query(query_idx, format, iterations, || {
                     let (row_count, timing) = execute(&mut ctx, query_idx, query.as_str())
                         .unwrap_or_else(|err| {
@@ -384,6 +385,8 @@ impl SqlBenchmarkRunner {
                 let mut runs = Vec::with_capacity(iterations);
                 let mut result = None;
 
+                tracing::debug!(%format, query_idx, "Running query");
+
                 for _ in 0..iterations {
                     let start = Instant::now();
                     let (row_count, timing, iter_result) = execute(&ctx, query_idx, query.as_str())

vortex-bench/src/tpch/benchmark.rs

@@ -18,7 +18,7 @@ use crate::tpch::EXPECTED_ROW_COUNTS_SF10;
 use crate::tpch::schema::CUSTOMER;
 use crate::tpch::schema::LINEITEM;
 use crate::tpch::schema::NATION;
-use crate::tpch::schema::ORDER;
+use crate::tpch::schema::ORDERS;
 use crate::tpch::schema::PART;
 use crate::tpch::schema::PARTSUPP;
 use crate::tpch::schema::REGION;
@@ -131,7 +131,7 @@ impl Benchmark for TpcHBenchmark {
             TableSpec::new("customer", Some(CUSTOMER.clone())),
             TableSpec::new("lineitem", Some(LINEITEM.clone())),
             TableSpec::new("nation", Some(NATION.clone())),
-            TableSpec::new("orders", Some(ORDER.clone())),
+            TableSpec::new("orders", Some(ORDERS.clone())),
             TableSpec::new("part", Some(PART.clone())),
             TableSpec::new("partsupp", Some(PARTSUPP.clone())),
             TableSpec::new("region", Some(REGION.clone())),

vortex-bench/src/tpch/schema.rs

@@ -82,7 +82,7 @@ pub static CUSTOMER: LazyLock<Schema> = LazyLock::new(|| {
     ])
 });
 
-pub static ORDER: LazyLock<Schema> = LazyLock::new(|| {
+pub static ORDERS: LazyLock<Schema> = LazyLock::new(|| {
     Schema::new(vec![
         Field::new("o_orderkey", DataType::Int64, false),
         Field::new("o_custkey", DataType::Int64, false),