Fix recalculation of skip indexes in ALTER UPDATE queries when table has adaptive granularity

src/Interpreters/MutationsInterpreter.cpp

@@ -40,6 +40,7 @@
 #include <Interpreters/InterpreterSelectQueryAnalyzer.h>
 #include <Parsers/makeASTForLogicalFunction.h>
 #include <Common/logger_useful.h>
+#include <Storages/MergeTree/MergeTreeDataPartType.h>
 
 namespace DB
 {
@@ -304,6 +305,11 @@ bool MutationsInterpreter::Source::hasProjection(const String & name) const
     return part && part->hasProjection(name);
 }
 
+bool MutationsInterpreter::Source::isCompactPart() const
+{
+    return part && part->getType() == MergeTreeDataPartType::Compact;
+}
+
 static Names getAvailableColumnsWithVirtuals(StorageMetadataPtr metadata_snapshot, const IStorage & storage)
 {
     auto all_columns = metadata_snapshot->getColumns().getNamesOfPhysical();
@@ -562,7 +568,8 @@ void MutationsInterpreter::prepare(bool dry_run)
     if (settings.recalculate_dependencies_of_updated_columns)
         dependencies = getAllColumnDependencies(metadata_snapshot, updated_columns, has_dependency);
 
-    bool has_alter_delete = false;
+    bool need_rebuild_indexes = false;
+    bool need_rebuild_projections = false;
     std::vector<String> read_columns;
 
     /// First, break a sequence of commands into stages.
@@ -583,7 +590,8 @@ void MutationsInterpreter::prepare(bool dry_run)
                 predicate = makeASTFunction("isZeroOrNull", predicate);
 
             stages.back().filters.push_back(predicate);
-            has_alter_delete = true;
+            need_rebuild_indexes = true;
+            need_rebuild_projections = true;
         }
         else if (command.type == MutationCommand::UPDATE)
         {
@@ -687,6 +695,9 @@ void MutationsInterpreter::prepare(bool dry_run)
                     }
                 }
             }
+
+            if (source.isCompactPart() && source.getMergeTreeData() && source.getMergeTreeData()->getSettings()->index_granularity_bytes > 0)
+                need_rebuild_indexes = true;
         }
         else if (command.type == MutationCommand::MATERIALIZE_COLUMN)
         {
@@ -892,7 +903,7 @@ void MutationsInterpreter::prepare(bool dry_run)
         if (!source.hasSecondaryIndex(index.name))
             continue;
 
-        if (has_alter_delete)
+        if (need_rebuild_indexes)
         {
             materialized_indices.insert(index.name);
             continue;
@@ -913,7 +924,7 @@ void MutationsInterpreter::prepare(bool dry_run)
         if (!source.hasProjection(projection.name))
             continue;
 
-        if (has_alter_delete)
+        if (need_rebuild_projections)
         {
             materialized_projections.insert(projection.name);
             continue;

src/Interpreters/MutationsInterpreter.h

@@ -122,6 +122,7 @@ class MutationsInterpreter
         bool materializeTTLRecalculateOnly() const;
         bool hasSecondaryIndex(const String & name) const;
         bool hasProjection(const String & name) const;
+        bool isCompactPart() const;
 
         void read(
             Stage & first_stage,

src/Storages/MergeTree/MergeTreeMarksLoader.cpp

@@ -107,21 +107,22 @@ MarkCache::MappedPtr MergeTreeMarksLoader::loadMarksImpl()
     // We first read the marks into a temporary simple array, then compress them into a more compact
     // representation.
     PODArray<MarkInCompressedFile> plain_marks(marks_count * columns_in_mark); // temporary
+    auto full_mark_path = std::string(fs::path(data_part_storage->getFullPath()) / mrk_path);
 
     if (file_size == 0 && marks_count != 0)
     {
         throw Exception(
             ErrorCodes::CORRUPTED_DATA,
             "Empty marks file '{}': {}, must be: {}",
-            std::string(fs::path(data_part_storage->getFullPath()) / mrk_path),
+            full_mark_path,
             file_size, expected_uncompressed_size);
     }
 
     if (!index_granularity_info.mark_type.compressed && expected_uncompressed_size != file_size)
         throw Exception(
             ErrorCodes::CORRUPTED_DATA,
             "Bad size of marks file '{}': {}, must be: {}",
-            std::string(fs::path(data_part_storage->getFullPath()) / mrk_path),
+            full_mark_path,
             file_size,
             expected_uncompressed_size);
 
@@ -142,7 +143,7 @@ MarkCache::MappedPtr MergeTreeMarksLoader::loadMarksImpl()
             throw Exception(
                 ErrorCodes::CANNOT_READ_ALL_DATA,
                 "Cannot read all marks from file {}, is eof: {}, buffer size: {}, file size: {}",
-                mrk_path,
+                full_mark_path,
                 reader->eof(),
                 reader->buffer().size(),
                 file_size);
@@ -155,7 +156,7 @@ MarkCache::MappedPtr MergeTreeMarksLoader::loadMarksImpl()
                 throw Exception(
                     ErrorCodes::CANNOT_READ_ALL_DATA,
                     "Cannot read all marks from file {}, marks expected {} (bytes size {}), marks read {} (bytes size {})",
-                    mrk_path, marks_count, expected_uncompressed_size, i, reader->count());
+                    full_mark_path, marks_count, expected_uncompressed_size, i, reader->count());
 
             size_t granularity;
             reader->readStrict(
@@ -167,7 +168,7 @@ MarkCache::MappedPtr MergeTreeMarksLoader::loadMarksImpl()
             throw Exception(
                 ErrorCodes::CANNOT_READ_ALL_DATA,
                 "Too many marks in file {}, marks expected {} (bytes size {})",
-                mrk_path, marks_count, expected_uncompressed_size);
+                full_mark_path, marks_count, expected_uncompressed_size);
     }
 
 #if __BYTE_ORDER__ == __ORDER_BIG_ENDIAN__

tests/queries/0_stateless/02891_alter_update_adaptive_granularity.reference

@@ -0,0 +1 @@
+342	442	The Containers library is a generic collection of class templates and algorithms that allow programmers to easily implement common data structures like queues, lists and stacks

tests/queries/0_stateless/02891_alter_update_adaptive_granularity.sql

@@ -0,0 +1,24 @@
+CREATE TABLE kv
+(
+    `key` UInt64,
+    `value` UInt64,
+    `s` String,
+    INDEX value_idx value TYPE minmax GRANULARITY 1
+)
+ENGINE = ReplacingMergeTree
+ORDER BY key
+SETTINGS index_granularity = 32, index_granularity_bytes = 1024;
+
+INSERT INTO kv SELECT
+    number,
+    number + 100,
+    toString(number)
+FROM numbers(2048);
+
+ALTER TABLE kv
+    UPDATE s = 'The Containers library is a generic collection of class templates and algorithms that allow programmers to easily implement common data structures like queues, lists and stacks' WHERE 1
+SETTINGS mutations_sync = 2;
+
+SELECT *
+FROM kv
+WHERE value = 442;