Merge pull request #974 from Altinity/backport/antalya-25.6.5/82304

Antalya 25.6: backport of ClickHouse#82304 - Better exception for non existing tables for data lakes

Author: Enmk

src/Databases/DataLake/DatabaseDataLake.cpp

@@ -68,6 +68,7 @@ namespace ErrorCodes
     extern const int BAD_ARGUMENTS;
     extern const int SUPPORT_IS_DISABLED;
     extern const int DATALAKE_DATABASE_ERROR;
+    extern const int CANNOT_GET_CREATE_TABLE_QUERY;
 }
 
 namespace
@@ -528,7 +529,12 @@ ASTPtr DatabaseDataLake::getCreateTableQueryImpl(
     auto table_metadata = DataLake::TableMetadata().withLocation().withSchema();
 
     const auto [namespace_name, table_name] = parseTableName(name);
-    catalog->getTableMetadata(namespace_name, table_name, table_metadata);
+
+    if (!catalog->tryGetTableMetadata(namespace_name, table_name, table_metadata))
+    {
+        throw Exception(
+            ErrorCodes::CANNOT_GET_CREATE_TABLE_QUERY, "Table `{}` doesn't exist", name);
+    }
 
     auto create_table_query = std::make_shared<ASTCreateQuery>();
     auto table_storage_define = table_engine_definition->clone();

src/Databases/DataLake/GlueCatalog.cpp

@@ -242,15 +242,6 @@ bool GlueCatalog::tryGetTableMetadata(
     const std::string & database_name,
     const std::string & table_name,
     TableMetadata & result) const
-{
-    getTableMetadata(database_name, table_name, result);
-    return true;
-}
-
-void GlueCatalog::getTableMetadata(
-    const std::string & database_name,
-    const std::string & table_name,
-    TableMetadata & result) const
 {
     Aws::Glue::Model::GetTableRequest request;
     request.SetDatabaseName(database_name);
@@ -320,11 +311,30 @@ void GlueCatalog::getTableMetadata(
     }
     else
     {
+        if (outcome.GetError().GetErrorType() == Aws::Glue::GlueErrors::ENTITY_NOT_FOUND)
+            return false; // Table does not exist
+
         throw DB::Exception(
             DB::ErrorCodes::DATALAKE_DATABASE_ERROR,
             "Exception calling GetTable for table {}: {}",
             database_name + "." + table_name, outcome.GetError().GetMessage());
     }
+
+    return true;
+}
+
+void GlueCatalog::getTableMetadata(
+    const std::string & database_name,
+    const std::string & table_name,
+    TableMetadata & result) const
+{
+    if (!tryGetTableMetadata(database_name, table_name, result))
+    {
+        throw DB::Exception(
+            DB::ErrorCodes::DATALAKE_DATABASE_ERROR,
+            "Table {} does not exist in Glue catalog",
+            database_name + "." + table_name);
+    }
 }
 
 void GlueCatalog::setCredentials(TableMetadata & metadata) const

tests/integration/test_database_glue/test.py

@@ -385,3 +385,47 @@ def test_select_after_rename(started_cluster):
             update.rename_column("bid", "new_bid")
 
         print(node.query(f"SELECT * FROM {CATALOG_NAME}.`{namespace}.{table_name}`"))
+
+
+def test_non_existing_tables(started_cluster):
+    node = started_cluster.instances["node1"]
+
+    test_ref = f"test_non_existing_tables_{uuid.uuid4()}"
+    table_name = f"{test_ref}_table"
+    root_namespace = f"{test_ref}_namespace"
+
+    namespaces_to_create = [
+        root_namespace,
+        f"{root_namespace}_A",
+    ]
+
+    catalog = load_catalog_impl(started_cluster)
+
+    for namespace in namespaces_to_create:
+        catalog.create_namespace(namespace)
+
+    for namespace in namespaces_to_create:
+        table = create_table(catalog, namespace, table_name)
+
+        num_rows = 10
+        df = generate_arrow_data(num_rows)
+        table.append(df)
+
+        create_clickhouse_glue_database(started_cluster, node, CATALOG_NAME)
+
+        expected = DEFAULT_CREATE_TABLE.format(CATALOG_NAME, namespace, table_name)
+        assert expected == node.query(
+            f"SHOW CREATE TABLE {CATALOG_NAME}.`{namespace}.{table_name}`"
+        )
+
+        try:
+            node.query(f"SHOW CREATE TABLE {CATALOG_NAME}.`{namespace}.wrong_table_name`")
+        except Exception as e:
+            assert "DB::Exception: Table" in str(e)
+            assert "doesn't exist" in str(e)
+
+        try:
+            node.query(f"SHOW CREATE TABLE {CATALOG_NAME}.`fake_namespace.wrong_table_name`")
+        except Exception as e:
+            assert "DB::Exception: Table" in str(e)
+            assert "doesn't exist" in str(e)

tests/integration/test_database_iceberg/test.py

@@ -373,3 +373,55 @@ def record(key):
 
     assert 'aaa\naaa\naaa' == node.query(f"SELECT symbol FROM {CATALOG_NAME}.`{namespace}.{table_name}`").strip()
     assert 'bbb\nbbb\nbbb' == node.query(f"SELECT symbol FROM {CATALOG_NAME}.`{namespace}.{table_name_2}`").strip()
+
+
+def test_non_existing_tables(started_cluster):
+    node = started_cluster.instances["node1"]
+
+    test_ref = f"test_list_tables_{uuid.uuid4()}"
+    table_name = f"{test_ref}_table"
+    root_namespace = f"{test_ref}_namespace"
+
+    namespace = f"{root_namespace}.A.B.C"
+    namespaces_to_create = [
+        root_namespace,
+        f"{root_namespace}.A",
+        f"{root_namespace}.A.B",
+        f"{root_namespace}.A.B.C",
+    ]
+
+    catalog = load_catalog_impl(started_cluster)
+
+    for namespace in namespaces_to_create:
+        catalog.create_namespace(namespace)
+        assert len(catalog.list_tables(namespace)) == 0
+
+    table = create_table(catalog, namespace, table_name)
+
+    num_rows = 10
+    data = [generate_record() for _ in range(num_rows)]
+    df = pa.Table.from_pylist(data)
+    table.append(df)
+
+    create_clickhouse_iceberg_database(started_cluster, node, CATALOG_NAME)
+
+    expected = DEFAULT_CREATE_TABLE.format(CATALOG_NAME, namespace, table_name)
+    assert expected == node.query(
+        f"SHOW CREATE TABLE {CATALOG_NAME}.`{namespace}.{table_name}`"
+    )
+
+    try:
+        node.query(
+            f"SHOW CREATE TABLE {CATALOG_NAME}.`{namespace}.qweqwe`"
+        )
+    except Exception as e:
+        assert "DB::Exception: Table" in str(e)
+        assert "doesn't exist" in str(e)
+
+    try:
+        node.query(
+            f"SHOW CREATE TABLE {CATALOG_NAME}.`qweqwe.qweqwe`"
+        )
+    except Exception as e:
+        assert "DB::Exception: Table" in str(e)
+        assert "doesn't exist" in str(e)