Merge pull request #7606 from xudong963/cbo_switch

chore: add the switch of enable_cbo

Author: mergify[bot]

src/query/service/src/sql/planner/optimizer/cascades/mod.rs

@@ -17,7 +17,9 @@ mod implement_rules;
 
 use std::collections::hash_map::Entry;
 use std::collections::HashMap;
+use std::sync::Arc;
 
+use common_catalog::table_context::TableContext;
 use common_exception::ErrorCode;
 use common_exception::Result;
 
@@ -47,17 +49,24 @@ pub struct CascadesOptimizer {
 
     /// group index -> best cost context
     best_cost_map: HashMap<IndexType, CostContext>,
+    _ctx: Arc<dyn TableContext>,
 }
 
 impl CascadesOptimizer {
-    pub fn create() -> Self {
-        CascadesOptimizer {
+    pub fn create(ctx: Arc<dyn TableContext>) -> Result<Self> {
+        let explore_rules = if ctx.get_settings().get_enable_cbo()? {
+            get_explore_rule_set()
+        } else {
+            RuleSet::create_with_ids(vec![]).unwrap()
+        };
+        Ok(CascadesOptimizer {
             memo: Memo::create(),
-            explore_rules: get_explore_rule_set(),
+            explore_rules,
             implement_rules: get_implement_rule_set(),
             cost_model: Box::new(DefaultCostModel),
             best_cost_map: HashMap::new(),
-        }
+            _ctx: ctx,
+        })
     }
 
     fn init(&mut self, expression: SExpr) -> Result<()> {

src/query/service/src/sql/planner/optimizer/mod.rs

@@ -140,10 +140,10 @@ pub fn optimize_query(
 
     let contains_local_table_scan = contains_local_table_scan(&s_expr, &metadata);
 
-    let mut heuristic = HeuristicOptimizer::new(ctx, bind_context, metadata, rules);
+    let mut heuristic = HeuristicOptimizer::new(ctx.clone(), bind_context, metadata, rules);
     let mut result = heuristic.optimize(s_expr)?;
 
-    let cascades = CascadesOptimizer::create();
+    let cascades = CascadesOptimizer::create(ctx)?;
     result = cascades.optimize(result)?;
 
     // So far, we don't have ability to execute distributed query

src/query/service/tests/it/storages/system/settings_table.rs

@@ -37,6 +37,7 @@ async fn test_settings_table() -> Result<()> {
         "| compression                    | None       | None       | SESSION | Format compression, default value: None                                                            | String |",
         "| empty_as_default               | 1          | 1          | SESSION | Format empty_as_default, default value: 1                                                          | UInt64 |",
         "| enable_async_insert            | 0          | 0          | SESSION | Whether the client open async insert mode, default value: 0                                        | UInt64 |",
+        "| enable_cbo                     | 1          | 1          | SESSION | If enable cost based optimization, default value: 1                                                | UInt64 |",
         "| enable_new_processor_framework | 1          | 1          | SESSION | Enable new processor framework if value != 0, default value: 1                                     | UInt64 |",
         "| enable_planner_v2              | 1          | 1          | SESSION | Enable planner v2 by setting this variable to 1, default value: 1                                  | UInt64 |",
         "| field_delimiter                | ,          | ,          | SESSION | Format field delimiter, default value: ,                                                           | String |",

src/query/settings/src/lib.rs

@@ -281,6 +281,13 @@ impl Settings {
                 desc: "SQL dialect, support \"PostgreSQL\" and \"MySQL\", default value: \"PostgreSQL\"",
                 possible_values: Some(vec!["PostgreSQL", "MySQL"]),
             },
+            SettingValue {
+                default_value: DataValue::UInt64(1),
+                user_setting: UserSetting::create("enable_cbo", DataValue::UInt64(1)),
+                level: ScopeLevel::Session,
+                desc: "If enable cost based optimization, default value: 1",
+                possible_values: None,
+            },
             // max_execute_time
             SettingValue {
                 default_value: DataValue::UInt64(0),
@@ -459,6 +466,18 @@ impl Settings {
         self.try_set_u64(KEY, v, false)
     }
 
+    pub fn get_enable_cbo(&self) -> Result<bool> {
+        static KEY: &str = "enable_cbo";
+        let v = self.try_get_u64(KEY)?;
+        Ok(v != 0)
+    }
+
+    pub fn set_enable_cbo(&self, val: bool) -> Result<()> {
+        static KEY: &str = "enable_cbo";
+        let v = if val { 1 } else { 0 };
+        self.try_set_u64(KEY, v, false)
+    }
+
     pub fn get_sql_dialect(&self) -> Result<Dialect> {
         let key = "sql_dialect";
         self.check_and_get_setting_value(key)

tests/logictest/suites/base/06_show/06_0003_show_settings

@@ -11,6 +11,7 @@ SHOW SETTINGS;
 compression  None  None  SESSION  Format compression, default value: None  String
 empty_as_default  1  1  SESSION  Format empty_as_default, default value: 1  UInt64
 enable_async_insert  0  0  SESSION  Whether the client open async insert mode, default value: 0  UInt64
+enable_cbo  1  1  SESSION  If enable cost based optimization, default value: 1  UInt64
 enable_new_processor_framework  1  1  SESSION  Enable new processor framework if value != 0, default value: 1  UInt64
 enable_planner_v2  1  1  SESSION  Enable planner v2 by setting this variable to 1, default value: 1  UInt64
 field_delimiter  ,  ,  SESSION  Format field delimiter, default value: ,  String
@@ -35,6 +36,7 @@ SHOW SETTINGS LIKE 'enable%';
 
 ----
 enable_async_insert 0 0 SESSION Whether the client open async insert mode, default value: 0 UInt64
+enable_cbo  1  1  SESSION  If enable cost based optimization, default value: 1  UInt64
 enable_new_processor_framework 1 1 SESSION Enable new processor framework if value != 0, default value: 1 UInt64
 enable_planner_v2 1 1 SESSION Enable planner v2 by setting this variable to 1, default value: 1 UInt64