Merge pull request #39181 from azat/part_log-merge_algorithm

Add merge_algorithm to system.part_log

docs/en/operations/system-tables/part_log.md

@@ -19,6 +19,10 @@ The `system.part_log` table contains the following columns:
     -   `REGULAR_MERGE` — Some regular merge.
     -   `TTL_DELETE_MERGE` — Cleaning up expired data.
     -   `TTL_RECOMPRESS_MERGE` — Recompressing data part with the.
+-   `merge_algorithm` ([Enum8](../../sql-reference/data-types/enum.md)) — Merge algorithm for the event with type `MERGE_PARTS`. Can have one of the following values:
+    -   `UNDECIDED`
+    -   `HORIZONTAL`
+    -   `VERTICAL`
 -   `event_date` ([Date](../../sql-reference/data-types/date.md)) — Event date.
 -   `event_time` ([DateTime](../../sql-reference/data-types/datetime.md)) — Event time.
 -   `event_time_microseconds` ([DateTime64](../../sql-reference/data-types/datetime64.md)) — Event time with microseconds precision.
@@ -52,6 +56,7 @@ Row 1:
 query_id:                      983ad9c7-28d5-4ae1-844e-603116b7de31
 event_type:                    NewPart
 merge_reason:                  NotAMerge
+merge_algorithm:               Undecided
 event_date:                    2021-02-02
 event_time:                    2021-02-02 11:14:28
 event_time_microseconds:                    2021-02-02 11:14:28.861919

docs/ru/operations/system-tables/part_log.md

@@ -14,6 +14,15 @@
     -   `REMOVE_PART` — удаление или отсоединение из таблицы с помощью [DETACH PARTITION](../../sql-reference/statements/alter/partition.md#alter_detach-partition).
     -   `MUTATE_PART` — изменение куска.
     -   `MOVE_PART` — перемещение куска между дисками.
+-   `merge_reason` ([Enum8](../../sql-reference/data-types/enum.md)) — Причина события с типом `MERGE_PARTS`. Может принимать одно из следующих значений:
+    -   `NOT_A_MERGE` — событие имеет тип иной, чем `MERGE_PARTS`.
+    -   `REGULAR_MERGE` — обычное слияние.
+    -   `TTL_DELETE_MERGE` — очистка истекших данных.
+    -   `TTL_RECOMPRESS_MERGE` — переупаковка куска.
+-   `merge_algorithm` ([Enum8](../../sql-reference/data-types/enum.md)) — Алгоритм слияния для события с типом `MERGE_PARTS`. Может принимать одно из следующих значений:
+    -   `UNDECIDED`
+    -   `HORIZONTAL`
+    -   `VERTICAL`
 -   `event_date` ([Date](../../sql-reference/data-types/date.md)) — дата события.
 -   `event_time` ([DateTime](../../sql-reference/data-types/datetime.md)) — время события.
 -   `event_time_microseconds` ([DateTime64](../../sql-reference/data-types/datetime64.md)) — время события с точностью до микросекунд.
@@ -46,6 +55,8 @@ Row 1:
 ──────
 query_id:                      983ad9c7-28d5-4ae1-844e-603116b7de31
 event_type:                    NewPart
+merge_reason:                  NotAMerge
+merge_algorithm:               Undecided
 event_date:                    2021-02-02
 event_time:                    2021-02-02 11:14:28
 event_time_microseconds:                    2021-02-02 11:14:28.861919

src/Interpreters/PartLog.cpp

@@ -36,6 +36,21 @@ PartLogElement::MergeReasonType PartLogElement::getMergeReasonType(MergeType mer
     throw Exception(ErrorCodes::NOT_IMPLEMENTED, "Unknown MergeType {}", static_cast<UInt64>(merge_type));
 }
 
+PartLogElement::PartMergeAlgorithm PartLogElement::getMergeAlgorithm(MergeAlgorithm merge_algorithm_)
+{
+    switch (merge_algorithm_)
+    {
+        case MergeAlgorithm::Undecided:
+            return UNDECIDED;
+        case MergeAlgorithm::Horizontal:
+            return HORIZONTAL;
+        case MergeAlgorithm::Vertical:
+            return VERTICAL;
+    }
+
+    throw Exception(ErrorCodes::NOT_IMPLEMENTED, "Unknown MergeAlgorithm {}", static_cast<UInt64>(merge_algorithm_));
+}
+
 NamesAndTypesList PartLogElement::getNamesAndTypes()
 {
     auto event_type_datatype = std::make_shared<DataTypeEnum8>(
@@ -60,12 +75,22 @@ NamesAndTypesList PartLogElement::getNamesAndTypes()
         }
     );
 
+    auto merge_algorithm_datatype = std::make_shared<DataTypeEnum8>(
+        DataTypeEnum8::Values
+        {
+            {"Undecided",  static_cast<Int8>(UNDECIDED)},
+            {"Horizontal", static_cast<Int8>(HORIZONTAL)},
+            {"Vertical",   static_cast<Int8>(VERTICAL)},
+        }
+    );
+
     ColumnsWithTypeAndName columns_with_type_and_name;
 
     return {
         {"query_id", std::make_shared<DataTypeString>()},
         {"event_type", std::move(event_type_datatype)},
         {"merge_reason", std::move(merge_reason_datatype)},
+        {"merge_algorithm", std::move(merge_algorithm_datatype)},
         {"event_date", std::make_shared<DataTypeDate>()},
 
         {"event_time", std::make_shared<DataTypeDateTime>()},
@@ -104,6 +129,7 @@ void PartLogElement::appendToBlock(MutableColumns & columns) const
     columns[i++]->insert(query_id);
     columns[i++]->insert(event_type);
     columns[i++]->insert(merge_reason);
+    columns[i++]->insert(merge_algorithm);
     columns[i++]->insert(DateLUT::instance().toDayNum(event_time).toUnderType());
     columns[i++]->insert(event_time);
     columns[i++]->insert(event_time_microseconds);

src/Interpreters/PartLog.h

@@ -5,6 +5,7 @@
 #include <Core/NamesAndTypes.h>
 #include <Core/NamesAndAliases.h>
 #include <Storages/MergeTree/MergeType.h>
+#include <Storages/MergeTree/MergeAlgorithm.h>
 
 
 namespace DB
@@ -22,6 +23,14 @@ struct PartLogElement
         MOVE_PART = 6,
     };
 
+    /// Copy of MergeAlgorithm since values are written to disk.
+    enum PartMergeAlgorithm
+    {
+        UNDECIDED = 0,
+        VERTICAL = 1,
+        HORIZONTAL = 2,
+    };
+
     enum MergeReasonType
     {
         /// merge_reason is relevant only for event_type = 'MERGE_PARTS', in other cases it is NOT_A_MERGE
@@ -38,6 +47,7 @@ struct PartLogElement
 
     Type event_type = NEW_PART;
     MergeReasonType merge_reason = NOT_A_MERGE;
+    PartMergeAlgorithm merge_algorithm = UNDECIDED;
 
     time_t event_time = 0;
     Decimal64 event_time_microseconds = 0;
@@ -72,6 +82,8 @@ struct PartLogElement
     static std::string name() { return "PartLog"; }
 
     static MergeReasonType getMergeReasonType(MergeType merge_type);
+    static PartMergeAlgorithm getMergeAlgorithm(MergeAlgorithm merge_algorithm_);
+
     static NamesAndTypesList getNamesAndTypes();
     static NamesAndAliases getNamesAndAliases() { return {}; }
     void appendToBlock(MutableColumns & columns) const;

src/Storages/MergeTree/MergeTreeData.cpp

@@ -6229,8 +6229,13 @@ try
     part_log_elem.event_type = type;
 
     if (part_log_elem.event_type == PartLogElement::MERGE_PARTS)
+    {
         if (merge_entry)
+        {
             part_log_elem.merge_reason = PartLogElement::getMergeReasonType((*merge_entry)->merge_type);
+            part_log_elem.merge_algorithm = PartLogElement::getMergeAlgorithm((*merge_entry)->merge_algorithm);
+        }
+    }
 
     part_log_elem.error = static_cast<UInt16>(execution_status.code);
     part_log_elem.exception = execution_status.message;

tests/queries/0_stateless/02362_part_log_merge_algorithm.reference

@@ -0,0 +1,5 @@
+data_horizontal	all_1_1_0	NewPart	Undecided
+data_horizontal	all_1_1_1	MergeParts	Horizontal
+data_vertical	all_1_1_0	NewPart	Undecided
+data_vertical	all_2_2_0	NewPart	Undecided
+data_vertical	all_1_2_1	MergeParts	Vertical

tests/queries/0_stateless/02362_part_log_merge_algorithm.sql

@@ -0,0 +1,26 @@
+CREATE TABLE data_horizontal (
+    key Int
+)
+Engine=MergeTree()
+ORDER BY key;
+
+INSERT INTO data_horizontal VALUES (1);
+OPTIMIZE TABLE data_horizontal FINAL;
+SYSTEM FLUSH LOGS;
+SELECT table, part_name, event_type, merge_algorithm FROM system.part_log WHERE event_date >= yesterday() AND database = currentDatabase() AND table = 'data_horizontal' ORDER BY event_time_microseconds;
+
+CREATE TABLE data_vertical
+(
+    key UInt64,
+    value String
+)
+ENGINE = MergeTree()
+ORDER BY key
+SETTINGS index_granularity_bytes = 0, enable_mixed_granularity_parts = 0, min_bytes_for_wide_part = 0,
+vertical_merge_algorithm_min_rows_to_activate = 1, vertical_merge_algorithm_min_columns_to_activate = 1;
+
+INSERT INTO data_vertical VALUES (1, '1');
+INSERT INTO data_vertical VALUES (2, '2');
+OPTIMIZE TABLE data_vertical FINAL;
+SYSTEM FLUSH LOGS;
+SELECT table, part_name, event_type, merge_algorithm FROM system.part_log WHERE event_date >= yesterday() AND database = currentDatabase() AND table = 'data_vertical' ORDER BY event_time_microseconds;