feat(benchmark): add generalized ASAP benchmark pipeline with ClickHouse and Elasticsearch support

asap-common/.gitignore

@@ -5,6 +5,7 @@
 .vscode/
 
 dependencies/py/promql_utilities/promql_utilities.egg-info/
+dependencies/py/promql_utilities/build/
 dependencies/rs/**/target/
 
 tests/**/*.json

asap-common/dependencies/rs/sql_utilities/src/ast_matching/sqlparser_test.rs

@@ -496,6 +496,65 @@ mod tests {
         );
     }
 
+    // ── ClickHouse parametric syntax + explicit BETWEEN timestamps ────────────
+    // These verify that a fully ClickHouse-compatible query (no DATEADD, no NOW())
+    // is parseable by ASAP: quantile(q)(col) + BETWEEN 'start' AND 'end'.
+
+    #[test]
+    fn test_clickhouse_explicit_datetime_temporal_quantile() {
+        check_query(
+            "SELECT quantile(0.95)(value) FROM cpu_usage WHERE time BETWEEN '2025-10-01 00:00:00' AND '2025-10-01 00:00:10' GROUP BY L1, L2, L3, L4",
+            vec![QueryType::TemporalQuantile],
+            None,
+        );
+    }
+
+    #[test]
+    // ASAP-only: parse_datetime accepts the Z suffix (interprets as UTC), but ClickHouse
+    // rejects it with TYPE_MISMATCH when comparing against a DateTime column.
+    // Do not use Z-suffix strings in queries intended for both systems.
+    fn test_asap_only_iso_z_temporal_quantile() {
+        check_query(
+            "SELECT quantile(0.95)(value) FROM cpu_usage WHERE time BETWEEN '2025-10-01T00:00:00Z' AND '2025-10-01T00:00:10Z' GROUP BY L1, L2, L3, L4",
+            vec![QueryType::TemporalQuantile],
+            None,
+        );
+    }
+
+    #[test]
+    // Both ASAP (parse_datetime) and ClickHouse treat ISO-without-Z as local server time.
+    // They agree only when running in the same timezone; prefer 'YYYY-MM-DD HH:MM:SS'
+    // (space format) to avoid this implicit dependency.
+    fn test_iso_no_z_treated_as_local_time_temporal_quantile() {
+        check_query(
+            "SELECT quantile(0.95)(value) FROM cpu_usage WHERE time BETWEEN '2025-10-01T00:00:00' AND '2025-10-01T00:00:10' GROUP BY L1, L2, L3, L4",
+            vec![QueryType::TemporalQuantile],
+            None,
+        );
+    }
+
+    #[test]
+    fn test_clickhouse_explicit_datetime_spatial_quantile() {
+        check_query(
+            "SELECT quantile(0.95)(value) FROM cpu_usage WHERE time BETWEEN '2025-10-01 00:00:00' AND '2025-10-01 00:00:01' GROUP BY L1",
+            vec![QueryType::Spatial],
+            None,
+        );
+    }
+
+    #[test]
+    fn test_clickhouse_explicit_matches_now_template() {
+        // A ClickHouse-style query (explicit timestamps, parametric quantile) must
+        // match a stored DATEADD(NOW()) template of the same shape.
+        let template = parse_sql_query(
+            "SELECT quantile(0.95)(value) FROM cpu_usage WHERE time BETWEEN DATEADD(s, -10, NOW()) AND NOW() GROUP BY L1, L2, L3, L4"
+        ).unwrap();
+        let incoming = parse_sql_query(
+            "SELECT quantile(0.95)(value) FROM cpu_usage WHERE time BETWEEN '2025-10-01 00:00:00' AND '2025-10-01 00:00:10' GROUP BY L1, L2, L3, L4"
+        ).unwrap();
+        assert!(incoming.matches_sql_pattern(&template));
+    }
+
     // ── Error cases ──────────────────────────────────────────────────────────
 
     #[test]

asap-common/dependencies/rs/sql_utilities/src/ast_matching/sqlpattern_parser.rs

@@ -325,6 +325,11 @@ impl SQLPatternParser {
     }
 
     fn get_timestamp_from_datetime_str(datetime_str: &str) -> Option<f64> {
+        // parse_datetime treats timezone-naive strings (e.g. "2025-10-01 00:00:00",
+        // "2025-10-01T00:00:00") as local server time, matching ClickHouse's behavior —
+        // but only when both run in the same timezone. Z-suffix strings (e.g.
+        // "2025-10-01T00:00:00Z") are interpreted as UTC here but rejected by ClickHouse.
+        // Use space-format datetime strings ("YYYY-MM-DD HH:MM:SS") for portability.
         let parsed_datetime = parse_datetime(datetime_str).ok()?;
         Some(parsed_datetime.timestamp().as_second() as f64)
     }

asap-query-engine/.gitignore

@@ -1 +1,2 @@
 target
+output/

asap-summary-ingest/.gitignore

@@ -1,3 +1,4 @@
 __pycache__
 **/*.pyc
 **/*.swp
+outputs/

asap-tools/execution-utilities/.gitignore

@@ -7,6 +7,9 @@
 
 clickhouse-benchmark-pipeline/benchmark_results/
 **/data/
+benchmark/arroyo_outputs/
+benchmark/queries/
+benchmark/results/
 
 **/*.csv
 **/*.png