Merge pull request ClickHouse#88917 from ClickHouse/vdimir/issue_88900

Forbid setting temporary_files_buffer_size=0

Author: PedroTadim

src/Core/Settings.cpp

@@ -3202,7 +3202,7 @@ Possible values:
 - NONE — No compression is applied.
 )", 0) \
     \
-    DECLARE(UInt64, temporary_files_buffer_size, DBMS_DEFAULT_BUFFER_SIZE, "Size of the buffer for temporary files writers. Larger buffer size means less system calls, but more memory consumption.", 0) \
+    DECLARE(NonZeroUInt64, temporary_files_buffer_size, DBMS_DEFAULT_BUFFER_SIZE, "Size of the buffer for temporary files writers. Larger buffer size means less system calls, but more memory consumption.", 0) \
     DECLARE(UInt64, max_rows_to_transfer, 0, R"(
 Maximum size (in rows) that can be passed to a remote server or saved in a
 temporary table when the GLOBAL IN/JOIN section is executed.

src/Interpreters/JoinOperator.cpp

@@ -49,7 +49,7 @@ namespace Setting
     extern const SettingsUInt64 max_joined_block_size_rows;
     extern const SettingsUInt64 max_joined_block_size_bytes;
     extern const SettingsString temporary_files_codec;
-    extern const SettingsUInt64 temporary_files_buffer_size;
+    extern const SettingsNonZeroUInt64 temporary_files_buffer_size;
     extern const SettingsUInt64 join_output_by_rowlist_perkey_rows_threshold;
     extern const SettingsUInt64 join_to_sort_minimum_perkey_rows;
     extern const SettingsUInt64 join_to_sort_maximum_table_rows;
@@ -92,7 +92,7 @@ namespace QueryPlanSerializationSetting
     extern const QueryPlanSerializationSettingsUInt64 max_joined_block_size_rows;
     extern const QueryPlanSerializationSettingsUInt64 max_joined_block_size_bytes;
     extern const QueryPlanSerializationSettingsString temporary_files_codec;
-    extern const QueryPlanSerializationSettingsUInt64 temporary_files_buffer_size;
+    extern const QueryPlanSerializationSettingsNonZeroUInt64 temporary_files_buffer_size;
     extern const QueryPlanSerializationSettingsUInt64 join_output_by_rowlist_perkey_rows_threshold;
     extern const QueryPlanSerializationSettingsUInt64 join_to_sort_minimum_perkey_rows;
     extern const QueryPlanSerializationSettingsUInt64 join_to_sort_maximum_table_rows;

src/Interpreters/ProcessList.cpp

@@ -50,7 +50,7 @@ namespace Setting
     extern const SettingsBool replace_running_query;
     extern const SettingsMilliseconds replace_running_query_max_wait_ms;
     extern const SettingsString temporary_files_codec;
-    extern const SettingsUInt64 temporary_files_buffer_size;
+    extern const SettingsNonZeroUInt64 temporary_files_buffer_size;
     extern const SettingsOverflowMode timeout_overflow_mode;
     extern const SettingsBool trace_profile_events;
     extern const SettingsMilliseconds low_priority_query_wait_time_ms;

src/Processors/QueryPlan/QueryPlanSerializationSettings.cpp

@@ -77,7 +77,7 @@ namespace DB
     \
     DECLARE(UInt64, max_rows_in_set_to_optimize_join, 0, "Maximal size of the set to filter joined tables by each other's row sets before joining.", 0) \
     DECLARE(String, temporary_files_codec, "LZ4", "Sets compression codec for temporary files used in sorting and joining operations on disk.", 0) \
-    DECLARE(UInt64, temporary_files_buffer_size, DBMS_DEFAULT_BUFFER_SIZE, "Size of the buffer for temporary files writers. Larger buffer size means less system calls, but more memory consumption.", 0) \
+    DECLARE(NonZeroUInt64, temporary_files_buffer_size, DBMS_DEFAULT_BUFFER_SIZE, "Size of the buffer for temporary files writers. Larger buffer size means less system calls, but more memory consumption.", 0) \
     \
     DECLARE(Bool, collect_hash_table_stats_during_joins, true, "Enable collecting hash table statistics to optimize memory allocation", 0) \
     DECLARE(UInt64, max_size_to_preallocate_for_joins, 1'000'000'000'000, "For how many elements it is allowed to preallocate space in all hash tables in total before join", 0) \

tests/clickhouse-test

@@ -1170,6 +1170,7 @@ class SettingsRandomizer:
         "use_skip_indexes_if_final": lambda: random.randint(0, 1),
         "use_skip_indexes_on_data_read": lambda: random.randint(0, 1),
         "optimize_rewrite_like_perfect_affix": lambda: random.randint(0, 1),
+        "temporary_files_buffer_size": lambda: random.randint(128, 2048),
         # Use the new reader most of the time.
         "input_format_parquet_use_native_reader_v3": lambda: min(1, random.randint(0, 5)),
         "enable_lazy_columns_replication": lambda: random.randint(0, 1),

tests/queries/0_stateless/03701_temporary_files_buffer_size.reference

@@ -0,0 +1,6 @@
+CREATE TABLE t0 (c0 Int) ENGINE = Memory();
+INSERT INTO TABLE t0 (c0) VALUES (1);
+SELECT c0 FROM t0 GROUP BY c0
+SETTINGS max_bytes_before_external_group_by = 1
+    , temporary_files_buffer_size = 0
+    , group_by_two_level_threshold = 1; -- { clientError BAD_ARGUMENTS }