Remove hive_statistics_merge_strategy from iceberg benchmarks (#43)

ZacBlanco · web-flow · commit 6f2154ec1765 · 2025-02-18T20:01:26.000-08:00
This is no longer required since Presto 289. The iceberg connector
natively supports storing data size and NDVs without use of the HMS,
even when the HMS is configured.

In fact, since HMS response caching was disabled, these parameters
negatively impact Iceberg query analysis performance. It is best to
leave the defaults.
diff --git a/benchmarks/tpc-ds/sf10k_ice.json b/benchmarks/tpc-ds/sf10k_ice.json
@@ -3,7 +3,6 @@
   "catalog": "iceberg",
   "schema": "tpcds_sf10000_parquet_iceberg",
   "session_params": {
-    "iceberg.hive_statistics_merge_strategy": "NUMBER_OF_DISTINCT_VALUES,TOTAL_SIZE_IN_BYTES",
     "optimizer_use_histograms": true
   }
 }
diff --git a/benchmarks/tpc-ds/sf10k_ice_par.json b/benchmarks/tpc-ds/sf10k_ice_par.json
@@ -3,7 +3,6 @@
   "catalog": "iceberg",
   "schema": "tpcds_sf10000_parquet_varchar_iceberg_part",
   "session_params": {
-    "iceberg.hive_statistics_merge_strategy": "NUMBER_OF_DISTINCT_VALUES,TOTAL_SIZE_IN_BYTES",
     "optimizer_use_histograms": true
   }
 }
diff --git a/benchmarks/tpc-ds/sf1k_ice.json b/benchmarks/tpc-ds/sf1k_ice.json
@@ -3,7 +3,6 @@
   "catalog": "iceberg",
   "schema": "tpcds_sf1000_parquet_iceberg",
   "session_params": {
-    "iceberg.hive_statistics_merge_strategy": "NUMBER_OF_DISTINCT_VALUES,TOTAL_SIZE_IN_BYTES",
     "optimizer_use_histograms": true
   }
 }
diff --git a/benchmarks/tpc-ds/sf1k_ice_par.json b/benchmarks/tpc-ds/sf1k_ice_par.json
@@ -3,7 +3,6 @@
   "catalog": "iceberg",
   "schema": "tpcds_sf1000_parquet_varchar_iceberg_part",
   "session_params": {
-    "iceberg.hive_statistics_merge_strategy": "NUMBER_OF_DISTINCT_VALUES,TOTAL_SIZE_IN_BYTES",
     "optimizer_use_histograms": true
   }
 }
diff --git a/benchmarks/tpc-ds/sf1k_ice_uncompressed.json b/benchmarks/tpc-ds/sf1k_ice_uncompressed.json
@@ -3,7 +3,6 @@
   "catalog": "iceberg",
   "schema": "tpcds_sf1000_parquet_uncompressed_iceberg",
   "session_params": {
-    "iceberg.hive_statistics_merge_strategy": "NUMBER_OF_DISTINCT_VALUES,TOTAL_SIZE_IN_BYTES",
     "optimizer_use_histograms": true
   }
 }
diff --git a/benchmarks/tpc-ds/sf30k_ice.json b/benchmarks/tpc-ds/sf30k_ice.json
@@ -3,7 +3,6 @@
   "catalog": "iceberg",
   "schema": "tpcds_sf30000_parquet_iceberg",
   "session_params": {
-    "iceberg.hive_statistics_merge_strategy": "NUMBER_OF_DISTINCT_VALUES,TOTAL_SIZE_IN_BYTES",
     "optimizer_use_histograms": true
   }
 }
diff --git a/benchmarks/tpch/sf1k_ice.json b/benchmarks/tpch/sf1k_ice.json
@@ -3,7 +3,6 @@
   "catalog": "iceberg",
   "schema": "tpch_sf1000_parquet_iceberg",
   "session_params": {
-    "iceberg.hive_statistics_merge_strategy": "NUMBER_OF_DISTINCT_VALUES,TOTAL_SIZE_IN_BYTES",
     "optimizer_use_histograms": true
   }
 }
diff --git a/benchmarks/tpch/sf1k_ice_par.json b/benchmarks/tpch/sf1k_ice_par.json
@@ -3,7 +3,6 @@
   "catalog": "iceberg",
   "schema": "tpch_sf1000_parquet_partitioned_iceberg",
   "session_params": {
-    "iceberg.hive_statistics_merge_strategy": "NUMBER_OF_DISTINCT_VALUES,TOTAL_SIZE_IN_BYTES",
     "optimizer_use_histograms": true
   }
 }

Original file line number	Diff line number	Diff line change
`@@ -3,7 +3,6 @@`
`3`	`3`	`"catalog": "iceberg",`
`4`	`4`	`"schema": "tpcds_sf10000_parquet_iceberg",`
`5`	`5`	`"session_params": {`
`6`		`- "iceberg.hive_statistics_merge_strategy": "NUMBER_OF_DISTINCT_VALUES,TOTAL_SIZE_IN_BYTES",`
`7`	`6`	`"optimizer_use_histograms": true`
`8`	`7`	`}`
`9`	`8`	`}`
Original file line number	Diff line number	Diff line change
`@@ -3,7 +3,6 @@`
`3`	`3`	`"catalog": "iceberg",`
`4`	`4`	`"schema": "tpcds_sf10000_parquet_varchar_iceberg_part",`
`5`	`5`	`"session_params": {`
`6`		`- "iceberg.hive_statistics_merge_strategy": "NUMBER_OF_DISTINCT_VALUES,TOTAL_SIZE_IN_BYTES",`
`7`	`6`	`"optimizer_use_histograms": true`
`8`	`7`	`}`
`9`	`8`	`}`
Original file line number	Diff line number	Diff line change
`@@ -3,7 +3,6 @@`
`3`	`3`	`"catalog": "iceberg",`
`4`	`4`	`"schema": "tpcds_sf1000_parquet_iceberg",`
`5`	`5`	`"session_params": {`
`6`		`- "iceberg.hive_statistics_merge_strategy": "NUMBER_OF_DISTINCT_VALUES,TOTAL_SIZE_IN_BYTES",`
`7`	`6`	`"optimizer_use_histograms": true`
`8`	`7`	`}`
`9`	`8`	`}`
Original file line number	Diff line number	Diff line change
`@@ -3,7 +3,6 @@`
`3`	`3`	`"catalog": "iceberg",`
`4`	`4`	`"schema": "tpcds_sf1000_parquet_varchar_iceberg_part",`
`5`	`5`	`"session_params": {`
`6`		`- "iceberg.hive_statistics_merge_strategy": "NUMBER_OF_DISTINCT_VALUES,TOTAL_SIZE_IN_BYTES",`
`7`	`6`	`"optimizer_use_histograms": true`
`8`	`7`	`}`
`9`	`8`	`}`
Original file line number	Diff line number	Diff line change
`@@ -3,7 +3,6 @@`
`3`	`3`	`"catalog": "iceberg",`
`4`	`4`	`"schema": "tpcds_sf1000_parquet_uncompressed_iceberg",`
`5`	`5`	`"session_params": {`
`6`		`- "iceberg.hive_statistics_merge_strategy": "NUMBER_OF_DISTINCT_VALUES,TOTAL_SIZE_IN_BYTES",`
`7`	`6`	`"optimizer_use_histograms": true`
`8`	`7`	`}`
`9`	`8`	`}`
Original file line number	Diff line number	Diff line change
`@@ -3,7 +3,6 @@`
`3`	`3`	`"catalog": "iceberg",`
`4`	`4`	`"schema": "tpcds_sf30000_parquet_iceberg",`
`5`	`5`	`"session_params": {`
`6`		`- "iceberg.hive_statistics_merge_strategy": "NUMBER_OF_DISTINCT_VALUES,TOTAL_SIZE_IN_BYTES",`
`7`	`6`	`"optimizer_use_histograms": true`
`8`	`7`	`}`
`9`	`8`	`}`
Original file line number	Diff line number	Diff line change
`@@ -3,7 +3,6 @@`
`3`	`3`	`"catalog": "iceberg",`
`4`	`4`	`"schema": "tpch_sf1000_parquet_iceberg",`
`5`	`5`	`"session_params": {`
`6`		`- "iceberg.hive_statistics_merge_strategy": "NUMBER_OF_DISTINCT_VALUES,TOTAL_SIZE_IN_BYTES",`
`7`	`6`	`"optimizer_use_histograms": true`
`8`	`7`	`}`
`9`	`8`	`}`
Original file line number	Diff line number	Diff line change
`@@ -3,7 +3,6 @@`
`3`	`3`	`"catalog": "iceberg",`
`4`	`4`	`"schema": "tpch_sf1000_parquet_partitioned_iceberg",`
`5`	`5`	`"session_params": {`
`6`		`- "iceberg.hive_statistics_merge_strategy": "NUMBER_OF_DISTINCT_VALUES,TOTAL_SIZE_IN_BYTES",`
`7`	`6`	`"optimizer_use_histograms": true`
`8`	`7`	`}`
`9`	`8`	`}`