Add server setting group_array_limit_size.

src/AggregateFunctions/AggregateFunctionGroupArray.cpp

@@ -753,13 +753,21 @@ size_t getMaxArraySize()
     return 0xFFFFFF;
 }
 
+bool isLimitArraySize()
+{
+    if (auto context = Context::getGlobalContextInstance())
+        return context->getServerSettings().aggregate_function_group_array_limit_size;
+
+    return false;
+}
+
 template <bool Tlast>
 AggregateFunctionPtr createAggregateFunctionGroupArray(
     const std::string & name, const DataTypes & argument_types, const Array & parameters, const Settings *)
 {
     assertUnary(name, argument_types);
 
-    bool limit_size = false;
+    bool limit_size = isLimitArraySize();
     UInt64 max_elems = getMaxArraySize();
 
     if (parameters.empty())

src/Core/ServerSettings.h

@@ -50,6 +50,7 @@ namespace DB
     M(UInt64, max_temporary_data_on_disk_size, 0, "The maximum amount of storage that could be used for external aggregation, joins or sorting., ", 0) \
     M(String, temporary_data_in_cache, "", "Cache disk name for temporary data.", 0) \
     M(UInt64, aggregate_function_group_array_max_element_size, 0xFFFFFF, "Max array element size in bytes for groupArray function. This limit is checked at serialization and help to avoid large state size.", 0) \
+    M(Bool, aggregate_function_group_array_limit_size, false, "This is set to true. Array elements are truncated when they exceed the max array element size.", 0) \
     M(UInt64, max_server_memory_usage, 0, "Maximum total memory usage of the server in bytes. Zero means unlimited.", 0) \
     M(Double, max_server_memory_usage_to_ram_ratio, 0.9, "Same as max_server_memory_usage but in to RAM ratio. Allows to lower max memory on low-memory systems.", 0) \
     M(UInt64, merges_mutations_memory_usage_soft_limit, 0, "Maximum total memory usage for merges and mutations in bytes. Zero means unlimited.", 0) \

tests/integration/test_group_array_element_size/configs/group_array_max_element_size.xml

@@ -1,3 +1,4 @@
 <clickhouse>
     <aggregate_function_group_array_max_element_size>10</aggregate_function_group_array_max_element_size>
+    <aggregate_function_group_array_limit_size>false</aggregate_function_group_array_limit_size>
 </clickhouse>

tests/integration/test_group_array_element_size/test.py

@@ -63,3 +63,32 @@ def test_max_exement_size(started_cluster):
     node1.restart_clickhouse()
 
     assert node1.query("select length(groupArrayMerge(x)) from tab3") == "21\n"
+
+def test_limit_size(started_cluster):
+    node1.query(
+        "CREATE TABLE tab4 (x AggregateFunction(groupArray, Array(UInt8))) ENGINE = MergeTree ORDER BY tuple()"
+    )
+    node1.query("insert into tab4 select groupArrayState([zero]) from zeros(10)")
+    assert node1.query("select length(groupArrayMerge(x)) from tab4") == "10\n"
+
+    node1.replace_in_config(
+        "/etc/clickhouse-server/config.d/group_array_max_element_size.xml",
+        "false",
+        "true",
+    )
+
+    node1.restart_clickhouse()
+
+    node1.query("insert into tab4 select groupArrayState([zero]) from zeros(100)")
+    assert node1.query("select length(groupArrayMerge(x)) from tab4") == "10\n"
+
+    node1.replace_in_config(
+        "/etc/clickhouse-server/config.d/group_array_max_element_size.xml",
+        "true",
+        "false",
+    )
+
+    node1.restart_clickhouse()
+
+    with pytest.raises(Exception, match=r"Too large array size"):
+        node1.query("insert into tab4 select groupArrayState([zero]) from zeros(11)")