diff --git a/.github/workflows/release-submodule.yaml b/.github/workflows/release-submodule.yaml
index 84ab2fc49a1c..af0ae78c32b2 100644
--- a/.github/workflows/release-submodule.yaml
+++ b/.github/workflows/release-submodule.yaml
@@ -75,9 +75,9 @@ jobs:
            package-name: ${{ matrix.package }}
            monorepo-tags: true
            command: release-pr
-      - id: label # Adds the "autorelease: pending" label.
-        if: ${{steps.release-please.outputs.pr}})
-        uses: actions/github-script@v3    
+      - uses: actions/github-script@v3
+        id: label # Adds the "autorelease: pending" label.
+        if: ${{steps.release-please.outputs.pr}}
         with:
             github-token: ${{secrets.GITHUB_TOKEN}}
             script: |
diff --git a/.github/workflows/release.yaml b/.github/workflows/release.yaml
index 34a176da17c6..d8489a725061 100644
--- a/.github/workflows/release.yaml
+++ b/.github/workflows/release.yaml
@@ -1,6 +1,7 @@
 # This workflow creates release PRs and tags the top level release for
 # the top level repository:
-push:
+on:
+  push:
     branches:
       - master
 name: release-please
@@ -17,9 +18,9 @@ jobs:
           fork: true
           package-name: google-cloud-go
           command: release-pr
-      - id: label # Add the magic "autorelease: pending" label.
+      - uses: actions/github-script@v3
+        id: label # Add the magic "autorelease: pending" label.
         if: ${{ steps.release-pr.outputs.pr }}
-        uses: actions/github-script@v3    
         with:
             github-token: ${{secrets.GITHUB_TOKEN}}
             script: |
diff --git a/README.md b/README.md
index fdf6f73c1c38..1ffe6700b5c5 100644
--- a/README.md
+++ b/README.md
@@ -175,3 +175,11 @@ for more information.
 [cloud-video]: https://cloud.google.com/video-intelligence/
 [cloud-vision]: https://cloud.google.com/vision
 [cloud-webrisk]: https://cloud.google.com/web-risk/
+
+## Links
+
+- [Go on Google Cloud](https://cloud.google.com/go/home)
+- [Getting started with Go on Google Cloud](https://cloud.google.com/go/getting-started)
+- [App Engine Quickstart](https://cloud.google.com/appengine/docs/standard/go/quickstart)
+- [Cloud Functions Quickstart](https://cloud.google.com/functions/docs/quickstart-go)
+- [Cloud Run Quickstart](https://cloud.google.com/run/docs/quickstarts/build-and-deploy#go)
diff --git a/bigquery/integration_test.go b/bigquery/integration_test.go
index e450e207ec5d..94d98f1401db 100644
--- a/bigquery/integration_test.go
+++ b/bigquery/integration_test.go
@@ -1630,6 +1630,63 @@ func TestIntegration_TableUpdate(t *testing.T) {
 	}
 }
 
+func TestIntegration_QueryStatistics(t *testing.T) {
+	// Make a bunch of assertions on a simple query.
+	if client == nil {
+		t.Skip("Integration tests skipped")
+	}
+	ctx := context.Background()
+
+	q := client.Query("SELECT 17 as foo, 3.14 as bar")
+	// disable cache to ensure we have query statistics
+	q.DisableQueryCache = true
+
+	job, err := q.Run(ctx)
+	if err != nil {
+		t.Fatalf("job Run failure: %v", err)
+	}
+	status, err := job.Wait(ctx)
+	if err != nil {
+		t.Fatalf("job Wait failure: %v", err)
+	}
+	if status.Statistics == nil {
+		t.Fatal("expected job statistics, none found")
+	}
+
+	if status.Statistics.NumChildJobs != 0 {
+		t.Errorf("expected no children, %d reported", status.Statistics.NumChildJobs)
+	}
+
+	if status.Statistics.ParentJobID != "" {
+		t.Errorf("expected no parent, but parent present: %s", status.Statistics.ParentJobID)
+	}
+
+	if status.Statistics.Details == nil {
+		t.Fatal("expected job details, none present")
+	}
+
+	qStats, ok := status.Statistics.Details.(*QueryStatistics)
+	if !ok {
+		t.Fatalf("expected query statistics not present")
+	}
+
+	if qStats.CacheHit {
+		t.Error("unexpected cache hit")
+	}
+
+	if qStats.StatementType != "SELECT" {
+		t.Errorf("expected SELECT statement type, got: %s", qStats.StatementType)
+	}
+
+	if len(qStats.QueryPlan) == 0 {
+		t.Error("expected query plan, none present")
+	}
+
+	if len(qStats.Timeline) == 0 {
+		t.Error("expected query timeline, none present")
+	}
+}
+
 func TestIntegration_Load(t *testing.T) {
 	if client == nil {
 		t.Skip("Integration tests skipped")
diff --git a/bigquery/job.go b/bigquery/job.go
index 6e0745d8d558..1e08e899c519 100644
--- a/bigquery/job.go
+++ b/bigquery/job.go
@@ -347,6 +347,9 @@ type JobStatistics struct {
 	// ScriptStatistics includes information run as part of a child job within
 	// a script.
 	ScriptStatistics *ScriptStatistics
+
+	// ReservationUsage attributes slot consumption to reservations.
+	ReservationUsage []*ReservationUsage
 }
 
 // Statistics is one of ExtractStatistics, LoadStatistics or QueryStatistics.
@@ -561,6 +564,25 @@ type QueryTimelineSample struct {
 	SlotMillis int64
 }
 
+// ReservationUsage contains information about a job's usage of a single reservation.
+type ReservationUsage struct {
+	// SlotMillis reports the slot milliseconds utilized within in the given reservation.
+	SlotMillis int64
+	// Name indicates the utilized reservation name, or "unreserved" for ondemand usage.
+	Name string
+}
+
+func bqToReservationUsage(ru []*bq.JobStatisticsReservationUsage) []*ReservationUsage {
+	var usage []*ReservationUsage
+	for _, in := range ru {
+		usage = append(usage, &ReservationUsage{
+			SlotMillis: in.SlotMs,
+			Name:       in.Name,
+		})
+	}
+	return usage
+}
+
 // ScriptStatistics report information about script-based query jobs.
 type ScriptStatistics struct {
 	EvaluationKind string
@@ -810,6 +832,7 @@ func (j *Job) setStatistics(s *bq.JobStatistics, c *Client) {
 		NumChildJobs:        s.NumChildJobs,
 		ParentJobID:         s.ParentJobId,
 		ScriptStatistics:    bqToScriptStatistics(s.ScriptStatistics),
+		ReservationUsage:    bqToReservationUsage(s.ReservationUsage),
 	}
 	switch {
 	case s.Extract != nil:
diff --git a/spanner/spannertest/README.md b/spanner/spannertest/README.md
index 5f4317e4cd92..bee71528e70f 100644
--- a/spanner/spannertest/README.md
+++ b/spanner/spannertest/README.md
@@ -18,6 +18,7 @@ Here's a list of features that are missing or incomplete. It is roughly ordered
 by ascending esotericism:
 
 - expression functions
+- NUMERIC
 - more aggregation functions
 - SELECT HAVING
 - case insensitivity
diff --git a/spanner/spansql/parser.go b/spanner/spansql/parser.go
index ce9badd8483e..9677c03bf282 100644
--- a/spanner/spansql/parser.go
+++ b/spanner/spansql/parser.go
@@ -1650,6 +1650,7 @@ var baseTypes = map[string]TypeBase{
 	"BOOL":      Bool,
 	"INT64":     Int64,
 	"FLOAT64":   Float64,
+	"NUMERIC":   Numeric,
 	"STRING":    String,
 	"BYTES":     Bytes,
 	"DATE":      Date,
@@ -1664,7 +1665,7 @@ func (p *parser) parseType() (Type, *parseError) {
 			ARRAY< scalar_type >
 
 		scalar_type:
-			{ BOOL | INT64 | FLOAT64 | STRING( length ) | BYTES( length ) | DATE | TIMESTAMP }
+			{ BOOL | INT64 | FLOAT64 | NUMERIC | STRING( length ) | BYTES( length ) | DATE | TIMESTAMP }
 		length:
 			{ int64_value | MAX }
 	*/
diff --git a/spanner/spansql/sql.go b/spanner/spansql/sql.go
index d5457de8676a..2c7cdcdf9f38 100644
--- a/spanner/spansql/sql.go
+++ b/spanner/spansql/sql.go
@@ -234,6 +234,8 @@ func (tb TypeBase) SQL() string {
 		return "INT64"
 	case Float64:
 		return "FLOAT64"
+	case Numeric:
+		return "NUMERIC"
 	case String:
 		return "STRING"
 	case Bytes:
diff --git a/spanner/spansql/types.go b/spanner/spansql/types.go
index 5854e0e0d64d..1c2831a07298 100644
--- a/spanner/spansql/types.go
+++ b/spanner/spansql/types.go
@@ -293,7 +293,7 @@ func (Check) isConstraint()   {}
 // Type represents a column type.
 type Type struct {
 	Array bool
-	Base  TypeBase // Bool, Int64, Float64, String, Bytes, Date, Timestamp
+	Base  TypeBase // Bool, Int64, Float64, Numeric, String, Bytes, Date, Timestamp
 	Len   int64    // if Base is String or Bytes; may be MaxLen
 }
 
@@ -306,6 +306,7 @@ const (
 	Bool TypeBase = iota
 	Int64
 	Float64
+	Numeric
 	String
 	Bytes
 	Date