TPC-DS: Schema and queries updated

Beuth-Erdelt · Nov 6, 2024 · af888be · af888be
1 parent 46efe6d
commit af888be
Show file tree

Hide file tree

Showing 5 changed files with 415 additions and 135 deletions.
diff --git a/docs/Example-TPC-DS.md b/docs/Example-TPC-DS.md
@@ -10,8 +10,17 @@ This example shows how to benchmark 99 reading queries Q1-Q99 derived from TPC-D
 > The query file is derived from the TPC-DS and as such is not comparable to published TPC-DS results, as the query file results do not comply with the TPC-DS Specification.
 
 In particular we had to apply changes, because
-* MySQL and MariaDB do not have a FULL OUTER JOIN
+* MySQL and MariaDB do not have a FULL OUTER JOIN (Q51, Q97)
 * MySQL and MariaDB do CASTing to INTEGER differently
+* column names may differ if AS is not used explicitly
+* MariaDB does not know GROUPING
+
+The schema of
+* MonetDB: sets primary keys and foreign key constraints after import and no other indexes
+* MySQL: sets primary keys before import, all foreign key constraints and indexes on foreign keys of fact tables and customer table after import
+* MariaDB: sets primary keys before import, all foreign key constraints and indexes on foreign keys of fact tables and customer table after import
+* PostgreSQL: sets primary keys before import, all foreign key constraints and indexes on foreign keys of fact tables and customer table after import
+
 
 See [query file](https://github.com/Beuth-Erdelt/Benchmark-Experiment-Host-Manager/blob/master/experiments/tpcds/queries-tpcds.config) for all details.
 

diff --git a/experiments/tpcds/MariaDB/initindexes-tpcds.sql b/experiments/tpcds/MariaDB/initindexes-tpcds.sql
@@ -1 +1,52 @@
 
+-- Store Sales Fact Table
+CREATE INDEX idx_store_sales_customer_id ON store_sales (ss_customer_sk);
+CREATE INDEX idx_store_sales_item_id ON store_sales (ss_item_sk);
+CREATE INDEX idx_store_sales_promo_id ON store_sales (ss_promo_sk);
+CREATE INDEX idx_store_sales_store_id ON store_sales (ss_store_sk);
+CREATE INDEX idx_store_sales_date_id ON store_sales (ss_sold_date_sk);
+
+-- Catalog Sales Fact Table
+CREATE INDEX idx_catalog_sales_customer_id ON catalog_sales (cs_bill_customer_sk);
+CREATE INDEX idx_catalog_sales_ship_customer_id ON catalog_sales (cs_ship_customer_sk);
+CREATE INDEX idx_catalog_sales_item_id ON catalog_sales (cs_item_sk);
+CREATE INDEX idx_catalog_sales_promo_id ON catalog_sales (cs_promo_sk);
+CREATE INDEX idx_catalog_sales_warehouse_id ON catalog_sales (cs_warehouse_sk);
+CREATE INDEX idx_catalog_sales_date_id ON catalog_sales (cs_sold_date_sk);
+
+-- Web Sales Fact Table
+CREATE INDEX idx_web_sales_customer_id ON web_sales (ws_bill_customer_sk);
+CREATE INDEX idx_web_sales_ship_customer_id ON web_sales (ws_ship_customer_sk);
+CREATE INDEX idx_web_sales_item_id ON web_sales (ws_item_sk);
+CREATE INDEX idx_web_sales_promo_id ON web_sales (ws_promo_sk);
+CREATE INDEX idx_web_sales_warehouse_id ON web_sales (ws_warehouse_sk);
+CREATE INDEX idx_web_sales_date_id ON web_sales (ws_sold_date_sk);
+
+-- Inventory Fact Table
+CREATE INDEX idx_inventory_item_id ON inventory (inv_item_sk);
+CREATE INDEX idx_inventory_warehouse_id ON inventory (inv_warehouse_sk);
+CREATE INDEX idx_inventory_date_id ON inventory (inv_date_sk);
+
+-- Customer Address Table (for customer relationships)
+CREATE INDEX idx_customer_address_id ON customer (c_current_addr_sk);
+CREATE INDEX idx_customer_birth_country ON customer (c_birth_country);
+
+-- Store Returns Table
+CREATE INDEX idx_store_returns_customer_id ON store_returns (sr_customer_sk);
+CREATE INDEX idx_store_returns_item_id ON store_returns (sr_item_sk);
+CREATE INDEX idx_store_returns_date_id ON store_returns (sr_returned_date_sk);
+CREATE INDEX idx_store_returns_ticket_id ON store_returns (sr_ticket_number);
+CREATE INDEX idx_store_returns_store_id ON store_returns (sr_store_sk);
+
+-- Catalog Returns Table
+CREATE INDEX idx_catalog_returns_customer_id ON catalog_returns (cr_returning_customer_sk);
+CREATE INDEX idx_catalog_returns_item_id ON catalog_returns (cr_item_sk);
+CREATE INDEX idx_catalog_returns_date_id ON catalog_returns (cr_returned_date_sk);
+CREATE INDEX idx_catalog_returns_order_number ON catalog_returns (cr_order_number);
+CREATE INDEX idx_catalog_returns_warehouse_id ON catalog_returns (cr_warehouse_sk);
+
+-- Web Returns Table
+CREATE INDEX idx_web_returns_customer_id ON web_returns (wr_returning_customer_sk);
+CREATE INDEX idx_web_returns_item_id ON web_returns (wr_item_sk);
+CREATE INDEX idx_web_returns_date_id ON web_returns (wr_returned_date_sk);
+CREATE INDEX idx_web_returns_order_number ON web_returns (wr_order_number);
diff --git a/experiments/tpcds/MySQL/initindexes-tpcds.sql b/experiments/tpcds/MySQL/initindexes-tpcds.sql
@@ -1 +1,52 @@
 
+-- Store Sales Fact Table
+CREATE INDEX idx_store_sales_customer_id ON store_sales (ss_customer_sk);
+CREATE INDEX idx_store_sales_item_id ON store_sales (ss_item_sk);
+CREATE INDEX idx_store_sales_promo_id ON store_sales (ss_promo_sk);
+CREATE INDEX idx_store_sales_store_id ON store_sales (ss_store_sk);
+CREATE INDEX idx_store_sales_date_id ON store_sales (ss_sold_date_sk);
+
+-- Catalog Sales Fact Table
+CREATE INDEX idx_catalog_sales_customer_id ON catalog_sales (cs_bill_customer_sk);
+CREATE INDEX idx_catalog_sales_ship_customer_id ON catalog_sales (cs_ship_customer_sk);
+CREATE INDEX idx_catalog_sales_item_id ON catalog_sales (cs_item_sk);
+CREATE INDEX idx_catalog_sales_promo_id ON catalog_sales (cs_promo_sk);
+CREATE INDEX idx_catalog_sales_warehouse_id ON catalog_sales (cs_warehouse_sk);
+CREATE INDEX idx_catalog_sales_date_id ON catalog_sales (cs_sold_date_sk);
+
+-- Web Sales Fact Table
+CREATE INDEX idx_web_sales_customer_id ON web_sales (ws_bill_customer_sk);
+CREATE INDEX idx_web_sales_ship_customer_id ON web_sales (ws_ship_customer_sk);
+CREATE INDEX idx_web_sales_item_id ON web_sales (ws_item_sk);
+CREATE INDEX idx_web_sales_promo_id ON web_sales (ws_promo_sk);
+CREATE INDEX idx_web_sales_warehouse_id ON web_sales (ws_warehouse_sk);
+CREATE INDEX idx_web_sales_date_id ON web_sales (ws_sold_date_sk);
+
+-- Inventory Fact Table
+CREATE INDEX idx_inventory_item_id ON inventory (inv_item_sk);
+CREATE INDEX idx_inventory_warehouse_id ON inventory (inv_warehouse_sk);
+CREATE INDEX idx_inventory_date_id ON inventory (inv_date_sk);
+
+-- Customer Address Table (for customer relationships)
+CREATE INDEX idx_customer_address_id ON customer (c_current_addr_sk);
+CREATE INDEX idx_customer_birth_country ON customer (c_birth_country);
+
+-- Store Returns Table
+CREATE INDEX idx_store_returns_customer_id ON store_returns (sr_customer_sk);
+CREATE INDEX idx_store_returns_item_id ON store_returns (sr_item_sk);
+CREATE INDEX idx_store_returns_date_id ON store_returns (sr_returned_date_sk);
+CREATE INDEX idx_store_returns_ticket_id ON store_returns (sr_ticket_number);
+CREATE INDEX idx_store_returns_store_id ON store_returns (sr_store_sk);
+
+-- Catalog Returns Table
+CREATE INDEX idx_catalog_returns_customer_id ON catalog_returns (cr_returning_customer_sk);
+CREATE INDEX idx_catalog_returns_item_id ON catalog_returns (cr_item_sk);
+CREATE INDEX idx_catalog_returns_date_id ON catalog_returns (cr_returned_date_sk);
+CREATE INDEX idx_catalog_returns_order_number ON catalog_returns (cr_order_number);
+CREATE INDEX idx_catalog_returns_warehouse_id ON catalog_returns (cr_warehouse_sk);
+
+-- Web Returns Table
+CREATE INDEX idx_web_returns_customer_id ON web_returns (wr_returning_customer_sk);
+CREATE INDEX idx_web_returns_item_id ON web_returns (wr_item_sk);
+CREATE INDEX idx_web_returns_date_id ON web_returns (wr_returned_date_sk);
+CREATE INDEX idx_web_returns_order_number ON web_returns (wr_order_number);
diff --git a/experiments/tpcds/PostgreSQL/initindexes-tpcds.sql b/experiments/tpcds/PostgreSQL/initindexes-tpcds.sql
@@ -1 +1,52 @@
 
+-- Store Sales Fact Table
+CREATE INDEX idx_store_sales_customer_id ON store_sales (ss_customer_sk);
+CREATE INDEX idx_store_sales_item_id ON store_sales (ss_item_sk);
+CREATE INDEX idx_store_sales_promo_id ON store_sales (ss_promo_sk);
+CREATE INDEX idx_store_sales_store_id ON store_sales (ss_store_sk);
+CREATE INDEX idx_store_sales_date_id ON store_sales (ss_sold_date_sk);
+
+-- Catalog Sales Fact Table
+CREATE INDEX idx_catalog_sales_customer_id ON catalog_sales (cs_bill_customer_sk);
+CREATE INDEX idx_catalog_sales_ship_customer_id ON catalog_sales (cs_ship_customer_sk);
+CREATE INDEX idx_catalog_sales_item_id ON catalog_sales (cs_item_sk);
+CREATE INDEX idx_catalog_sales_promo_id ON catalog_sales (cs_promo_sk);
+CREATE INDEX idx_catalog_sales_warehouse_id ON catalog_sales (cs_warehouse_sk);
+CREATE INDEX idx_catalog_sales_date_id ON catalog_sales (cs_sold_date_sk);
+
+-- Web Sales Fact Table
+CREATE INDEX idx_web_sales_customer_id ON web_sales (ws_bill_customer_sk);
+CREATE INDEX idx_web_sales_ship_customer_id ON web_sales (ws_ship_customer_sk);
+CREATE INDEX idx_web_sales_item_id ON web_sales (ws_item_sk);
+CREATE INDEX idx_web_sales_promo_id ON web_sales (ws_promo_sk);
+CREATE INDEX idx_web_sales_warehouse_id ON web_sales (ws_warehouse_sk);
+CREATE INDEX idx_web_sales_date_id ON web_sales (ws_sold_date_sk);
+
+-- Inventory Fact Table
+CREATE INDEX idx_inventory_item_id ON inventory (inv_item_sk);
+CREATE INDEX idx_inventory_warehouse_id ON inventory (inv_warehouse_sk);
+CREATE INDEX idx_inventory_date_id ON inventory (inv_date_sk);
+
+-- Customer Address Table (for customer relationships)
+CREATE INDEX idx_customer_address_id ON customer (c_current_addr_sk);
+CREATE INDEX idx_customer_birth_country ON customer (c_birth_country);
+
+-- Store Returns Table
+CREATE INDEX idx_store_returns_customer_id ON store_returns (sr_customer_sk);
+CREATE INDEX idx_store_returns_item_id ON store_returns (sr_item_sk);
+CREATE INDEX idx_store_returns_date_id ON store_returns (sr_returned_date_sk);
+CREATE INDEX idx_store_returns_ticket_id ON store_returns (sr_ticket_number);
+CREATE INDEX idx_store_returns_store_id ON store_returns (sr_store_sk);
+
+-- Catalog Returns Table
+CREATE INDEX idx_catalog_returns_customer_id ON catalog_returns (cr_returning_customer_sk);
+CREATE INDEX idx_catalog_returns_item_id ON catalog_returns (cr_item_sk);
+CREATE INDEX idx_catalog_returns_date_id ON catalog_returns (cr_returned_date_sk);
+CREATE INDEX idx_catalog_returns_order_number ON catalog_returns (cr_order_number);
+CREATE INDEX idx_catalog_returns_warehouse_id ON catalog_returns (cr_warehouse_sk);
+
+-- Web Returns Table
+CREATE INDEX idx_web_returns_customer_id ON web_returns (wr_returning_customer_sk);
+CREATE INDEX idx_web_returns_item_id ON web_returns (wr_item_sk);
+CREATE INDEX idx_web_returns_date_id ON web_returns (wr_returned_date_sk);
+CREATE INDEX idx_web_returns_order_number ON web_returns (wr_order_number);