Merge #60771

60771: sql,opt: add cluster and session settings for locality optimized search r=rytaft a=rytaft This commit adds a new session setting `locality_optimized_partitioned_index_scan` and corresponding cluster default setting `sql.defaults.locality_optimized_partitioned_index_scan.enabled`. In future commits, these settings will be used to enable or disable locality optimized search. If enabled, the optimizer will try to plan scans and lookup joins in which local nodes are searched for matching rows before remote nodes, in the hope that the execution engine can avoid visiting remote nodes. Informs #55185 Release note (sql change): Added a new session setting locality_optimized_partitioned_index_scan and corresponding cluster default setting sql.defaults.locality_optimized_partitioned_index_scan.enabled. Both are currently disabled by default, and are currently unused. In the future, these settings will be used to enable or disable locality optimized search. If enabled, the optimizer will try to search locally for rows in REGIONAL BY ROW tables before searching remote nodes. Co-authored-by: Rebecca Taft <[email protected]>
cockroachdb · Feb 20, 2021 · 3851e17 · 3851e17
2 parents 557346d + 1f724eb
commit 3851e17
Show file tree

Hide file tree

Showing 10 changed files with 56 additions and 0 deletions.
diff --git a/pkg/sql/exec_util.go b/pkg/sql/exec_util.go
@@ -287,6 +287,18 @@ var optUseMultiColStatsClusterMode = settings.RegisterBoolSetting(
 	true,
 )
 
+// localityOptimizedSearchMode controls the cluster default for the use of
+// locality optimized search. If enabled, the optimizer will try to plan scans
+// and lookup joins in which local nodes (i.e., nodes in the gateway region) are
+// searched for matching rows before remote nodes, in the hope that the
+// execution engine can avoid visiting remote nodes.
+var localityOptimizedSearchMode = settings.RegisterBoolSetting(
+	"sql.defaults.locality_optimized_partitioned_index_scan.enabled",
+	"default value for locality_optimized_partitioned_index_scan session setting; "+
+		"enables searching for rows in the current region before searching remote regions",
+	false,
+)
+
 var implicitSelectForUpdateClusterMode = settings.RegisterBoolSetting(
 	"sql.defaults.implicit_select_for_update.enabled",
 	"default value for enable_implicit_select_for_update session setting; enables FOR UPDATE locking during the row-fetch phase of mutation statements",
@@ -2195,6 +2207,10 @@ func (m *sessionDataMutator) SetOptimizerUseMultiColStats(val bool) {
 	m.data.OptimizerUseMultiColStats = val
 }
 
+func (m *sessionDataMutator) SetLocalityOptimizedSearch(val bool) {
+	m.data.LocalityOptimizedSearch = val
+}
+
 func (m *sessionDataMutator) SetImplicitSelectForUpdate(val bool) {
 	m.data.ImplicitSelectForUpdate = val
 }

diff --git a/pkg/sql/logictest/testdata/logic_test/information_schema b/pkg/sql/logictest/testdata/logic_test/information_schema
@@ -4191,6 +4191,7 @@ idle_in_transaction_session_timeout                   0
 integer_datetimes                                     on
 intervalstyle                                         postgres
 locality                                              region=test,dc=dc1
+locality_optimized_partitioned_index_scan             off
 lock_timeout                                          0
 max_identifier_length                                 128
 max_index_keys                                        32

diff --git a/pkg/sql/logictest/testdata/logic_test/pg_catalog b/pkg/sql/logictest/testdata/logic_test/pg_catalog
@@ -2456,6 +2456,7 @@ idle_in_transaction_session_timeout                   0                   NULL
 integer_datetimes                                     on                  NULL      NULL        NULL        string
 intervalstyle                                         postgres            NULL      NULL        NULL        string
 locality                                              region=test,dc=dc1  NULL      NULL        NULL        string
+locality_optimized_partitioned_index_scan             off                 NULL      NULL        NULL        string
 lock_timeout                                          0                   NULL      NULL        NULL        string
 max_identifier_length                                 128                 NULL      NULL        NULL        string
 max_index_keys                                        32                  NULL      NULL        NULL        string
@@ -2532,6 +2533,7 @@ idle_in_transaction_session_timeout                   0                   NULL
 integer_datetimes                                     on                  NULL  user     NULL      on                  on
 intervalstyle                                         postgres            NULL  user     NULL      postgres            postgres
 locality                                              region=test,dc=dc1  NULL  user     NULL      region=test,dc=dc1  region=test,dc=dc1
+locality_optimized_partitioned_index_scan             off                 NULL  user     NULL      off                 off
 lock_timeout                                          0                   NULL  user     NULL      0                   0
 max_identifier_length                                 128                 NULL  user     NULL      128                 128
 max_index_keys                                        32                  NULL  user     NULL      32                  32
@@ -2604,6 +2606,7 @@ idle_in_transaction_session_timeout                   NULL    NULL     NULL
 integer_datetimes                                     NULL    NULL     NULL     NULL        NULL
 intervalstyle                                         NULL    NULL     NULL     NULL        NULL
 locality                                              NULL    NULL     NULL     NULL        NULL
+locality_optimized_partitioned_index_scan             NULL    NULL     NULL     NULL        NULL
 lock_timeout                                          NULL    NULL     NULL     NULL        NULL
 max_identifier_length                                 NULL    NULL     NULL     NULL        NULL
 max_index_keys                                        NULL    NULL     NULL     NULL        NULL

diff --git a/pkg/sql/logictest/testdata/logic_test/show_source b/pkg/sql/logictest/testdata/logic_test/show_source
@@ -61,6 +61,7 @@ idle_in_transaction_session_timeout                   0
 integer_datetimes                                     on
 intervalstyle                                         postgres
 locality                                              region=test,dc=dc1
+locality_optimized_partitioned_index_scan             off
 lock_timeout                                          0
 max_identifier_length                                 128
 max_index_keys                                        32

diff --git a/pkg/sql/opt/memo/memo.go b/pkg/sql/opt/memo/memo.go
@@ -135,6 +135,7 @@ type Memo struct {
 	zigzagJoinEnabled       bool
 	useHistograms           bool
 	useMultiColStats        bool
+	localityOptimizedSearch bool
 	safeUpdates             bool
 	preferLookupJoinsForFKs bool
 	saveTablesPrefix        string
@@ -173,6 +174,7 @@ func (m *Memo) Init(evalCtx *tree.EvalContext) {
 		zigzagJoinEnabled:       evalCtx.SessionData.ZigzagJoinEnabled,
 		useHistograms:           evalCtx.SessionData.OptimizerUseHistograms,
 		useMultiColStats:        evalCtx.SessionData.OptimizerUseMultiColStats,
+		localityOptimizedSearch: evalCtx.SessionData.LocalityOptimizedSearch,
 		safeUpdates:             evalCtx.SessionData.SafeUpdates,
 		preferLookupJoinsForFKs: evalCtx.SessionData.PreferLookupJoinsForFKs,
 		saveTablesPrefix:        evalCtx.SessionData.SaveTablesPrefix,
@@ -281,6 +283,7 @@ func (m *Memo) IsStale(
 		m.zigzagJoinEnabled != evalCtx.SessionData.ZigzagJoinEnabled ||
 		m.useHistograms != evalCtx.SessionData.OptimizerUseHistograms ||
 		m.useMultiColStats != evalCtx.SessionData.OptimizerUseMultiColStats ||
+		m.localityOptimizedSearch != evalCtx.SessionData.LocalityOptimizedSearch ||
 		m.safeUpdates != evalCtx.SessionData.SafeUpdates ||
 		m.preferLookupJoinsForFKs != evalCtx.SessionData.PreferLookupJoinsForFKs ||
 		m.saveTablesPrefix != evalCtx.SessionData.SaveTablesPrefix {

diff --git a/pkg/sql/opt/memo/memo_test.go b/pkg/sql/opt/memo/memo_test.go
@@ -204,6 +204,12 @@ func TestMemoIsStale(t *testing.T) {
 	evalCtx.SessionData.OptimizerUseMultiColStats = false
 	notStale()
 
+	// Stale locality optimized search enable.
+	evalCtx.SessionData.LocalityOptimizedSearch = true
+	stale()
+	evalCtx.SessionData.LocalityOptimizedSearch = false
+	notStale()
+
 	// Stale safe updates.
 	evalCtx.SessionData.SafeUpdates = true
 	stale()

diff --git a/pkg/sql/opt/optbuilder/builder_test.go b/pkg/sql/opt/optbuilder/builder_test.go
@@ -78,6 +78,7 @@ func TestBuilder(t *testing.T) {
 				evalCtx := tree.MakeTestingEvalContext(cluster.MakeTestingClusterSettings())
 				evalCtx.SessionData.OptimizerUseHistograms = true
 				evalCtx.SessionData.OptimizerUseMultiColStats = true
+				evalCtx.SessionData.LocalityOptimizedSearch = true
 
 				var o xform.Optimizer
 				o.Init(&evalCtx, catalog)

diff --git a/pkg/sql/opt/testutils/opttester/opt_tester.go b/pkg/sql/opt/testutils/opttester/opt_tester.go
@@ -244,6 +244,7 @@ func New(catalog cat.Catalog, sql string) *OptTester {
 	ot.evalCtx.SessionData.ZigzagJoinEnabled = true
 	ot.evalCtx.SessionData.OptimizerUseHistograms = true
 	ot.evalCtx.SessionData.OptimizerUseMultiColStats = true
+	ot.evalCtx.SessionData.LocalityOptimizedSearch = true
 	ot.evalCtx.SessionData.ReorderJoinsLimit = opt.DefaultJoinOrderLimit
 	ot.evalCtx.SessionData.InsertFastPath = true
 

diff --git a/pkg/sql/sessiondata/session_data.go b/pkg/sql/sessiondata/session_data.go
@@ -178,6 +178,11 @@ type LocalOnlySessionData struct {
 	// OptimizerUseMultiColStats indicates whether we should use multi-column
 	// statistics for cardinality estimation in the optimizer.
 	OptimizerUseMultiColStats bool
+	// LocalityOptimizedSearch indicates that the optimizer will try to plan scans
+	// and lookup joins in which local nodes (i.e., nodes in the gateway region)
+	// are searched for matching rows before remote nodes, in the hope that the
+	// execution engine can avoid visiting remote nodes.
+	LocalityOptimizedSearch bool
 	// SafeUpdates causes errors when the client
 	// sends syntax that may have unwanted side effects.
 	SafeUpdates bool

diff --git a/pkg/sql/vars.go b/pkg/sql/vars.go
@@ -673,6 +673,25 @@ var varGen = map[string]sessionVar{
 		},
 	},
 
+	// CockroachDB extension.
+	`locality_optimized_partitioned_index_scan`: {
+		GetStringVal: makePostgresBoolGetStringValFn(`locality_optimized_partitioned_index_scan`),
+		Set: func(_ context.Context, m *sessionDataMutator, s string) error {
+			b, err := paramparse.ParseBoolVar(`locality_optimized_partitioned_index_scan`, s)
+			if err != nil {
+				return err
+			}
+			m.SetLocalityOptimizedSearch(b)
+			return nil
+		},
+		Get: func(evalCtx *extendedEvalContext) string {
+			return formatBoolAsPostgresSetting(evalCtx.SessionData.LocalityOptimizedSearch)
+		},
+		GlobalDefault: func(sv *settings.Values) string {
+			return formatBoolAsPostgresSetting(localityOptimizedSearchMode.Get(sv))
+		},
+	},
+
 	// CockroachDB extension.
 	`enable_implicit_select_for_update`: {
 		GetStringVal: makePostgresBoolGetStringValFn(`enable_implicit_select_for_update`),