Optimize mongodb collections Cutoff methods

cmd/api/main.go

@@ -12,6 +12,7 @@ import (
 	"github.com/aykhans/bsky-feedgen/pkg/logger"
 	"github.com/aykhans/bsky-feedgen/pkg/storage/mongodb"
 	"github.com/aykhans/bsky-feedgen/pkg/storage/mongodb/collections"
+	_ "go.uber.org/automaxprocs"
 )
 
 func main() {

cmd/consumer/main.go

@@ -17,6 +17,7 @@ import (
 	"github.com/aykhans/bsky-feedgen/pkg/logger"
 	"github.com/aykhans/bsky-feedgen/pkg/storage/mongodb"
 	"github.com/aykhans/bsky-feedgen/pkg/storage/mongodb/collections"
+	_ "go.uber.org/automaxprocs"
 )
 
 func main() {

cmd/feedgen/az/main.go

@@ -17,6 +17,7 @@ import (
 	"github.com/aykhans/bsky-feedgen/pkg/logger"
 	"github.com/aykhans/bsky-feedgen/pkg/storage/mongodb"
 	"github.com/aykhans/bsky-feedgen/pkg/storage/mongodb/collections"
+	_ "go.uber.org/automaxprocs"
 )
 
 func main() {

go.mod

@@ -8,6 +8,8 @@ require (
 	go.mongodb.org/mongo-driver v1.17.3
 )
 
+require go.uber.org/automaxprocs v1.6.0 // indirect
+
 require (
 	github.com/RussellLuo/slidingwindow v0.0.0-20200528002341-535bb99d338b // indirect
 	github.com/beorn7/perks v1.0.1 // indirect

go.sum

@@ -330,6 +330,8 @@ go.uber.org/atomic v1.6.0/go.mod h1:sABNBOSYdrvTF6hTgEIbc7YasKWGhgEQZyfxyTvoXHQ=
 go.uber.org/atomic v1.7.0/go.mod h1:fEN4uk6kAWBTFdckzkM89CLk9XfWZrxpCo0nPH17wJc=
 go.uber.org/atomic v1.11.0 h1:ZvwS0R+56ePWxUNi+Atn9dWONBPp/AUETXlHW0DxSjE=
 go.uber.org/atomic v1.11.0/go.mod h1:LUxbIzbOniOlMKjJjyPfpl4v+PKK2cNJn91OQbhoJI0=
+go.uber.org/automaxprocs v1.6.0 h1:O3y2/QNTOdbF+e/dpXNNW7Rx2hZ4sTIPyybbxyNqTUs=
+go.uber.org/automaxprocs v1.6.0/go.mod h1:ifeIMSnPZuznNm6jmdzmU3/bfk01Fe2fotchwEFJ8r8=
 go.uber.org/goleak v1.1.11-0.20210813005559-691160354723/go.mod h1:cwTWslyiVhfpKIDGSZEM2HlOvcqm+tG4zioyIeLoqMQ=
 go.uber.org/goleak v1.2.0 h1:xqgm/S+aQvhWFTtR0XK3Jvg7z8kGV8P4X14IzwN3Eqk=
 go.uber.org/goleak v1.2.0/go.mod h1:XJYK+MuIchqpmGmUSAzotztawfKvYLUIgg7guXrwVUo=

pkg/storage/mongodb/collections/feed_az.go

@@ -163,7 +163,8 @@ func (f FeedAzCollection) CutoffByCount(
 
 	findOpts := options.Find().
 		SetSort(bson.D{{Key: "created_at", Value: 1}}).
-		SetLimit(deleteCount)
+		SetLimit(deleteCount).
+		SetProjection(bson.M{"_id": 1})
 
 	cursor, err := f.Collection.Find(ctx, bson.M{}, findOpts)
 	if err != nil {
@@ -171,24 +172,46 @@ func (f FeedAzCollection) CutoffByCount(
 	}
 	defer func() { _ = cursor.Close(ctx) }()
 
-	var docsToDelete []bson.M
-	if err = cursor.All(ctx, &docsToDelete); err != nil {
-		return 0, err
+	// Process documents in batches to avoid potential memory issues
+	const batchSize = 10000
+	var totalDeleted int64 = 0
+
+	for {
+		batch := make([]string, 0, batchSize)
+		batchCount := 0
+
+		for cursor.Next(ctx) && batchCount < batchSize {
+			var doc struct {
+				ID string `bson:"_id"`
+			}
+			if err = cursor.Decode(&doc); err != nil {
+				return totalDeleted, err
+			}
+			batch = append(batch, doc.ID)
+			batchCount++
+		}
+
+		if len(batch) == 0 {
+			break
+		}
+
+		// Delete the batch
+		result, err := f.Collection.DeleteMany(ctx, bson.M{"_id": bson.M{"$in": batch}})
+		if err != nil {
+			return totalDeleted, err
+		}
+
+		totalDeleted += result.DeletedCount
+
+		if cursor.Err() != nil {
+			return totalDeleted, cursor.Err()
+		}
+
+		// If we didn't fill the batch, we're done
+		if batchCount < batchSize {
+			break
+		}
 	}
 
-	if len(docsToDelete) == 0 {
-		return 0, nil
-	}
-
-	ids := make([]any, len(docsToDelete))
-	for i := range docsToDelete {
-		ids[i] = docsToDelete[i]["_id"]
-	}
-
-	result, err := f.Collection.DeleteMany(ctx, bson.M{"_id": bson.M{"$in": ids}})
-	if err != nil {
-		return 0, err
-	}
-
-	return result.DeletedCount, nil
+	return totalDeleted, nil
 }

pkg/storage/mongodb/collections/post.go

@@ -17,10 +17,15 @@ type PostCollection struct {
 func NewPostCollection(client *mongo.Client) (*PostCollection, error) {
 	client.Database(config.MongoDBBaseDB).Collection("")
 	coll := client.Database(config.MongoDBBaseDB).Collection("post")
-	_, err := coll.Indexes().CreateOne(
+	_, err := coll.Indexes().CreateMany(
 		context.Background(),
-		mongo.IndexModel{
-			Keys: bson.D{{Key: "sequence", Value: -1}},
+		[]mongo.IndexModel{
+			{
+				Keys: bson.D{{Key: "sequence", Value: -1}},
+			},
+			{
+				Keys: bson.D{{Key: "created_at", Value: 1}},
+			},
 		},
 	)
 	if err != nil {
@@ -71,7 +76,8 @@ func (p PostCollection) CutoffByCount(
 
 	findOpts := options.Find().
 		SetSort(bson.D{{Key: "created_at", Value: 1}}).
-		SetLimit(deleteCount)
+		SetLimit(deleteCount).
+		SetProjection(bson.M{"_id": 1})
 
 	cursor, err := p.Collection.Find(ctx, bson.M{}, findOpts)
 	if err != nil {
@@ -79,26 +85,48 @@ func (p PostCollection) CutoffByCount(
 	}
 	defer func() { _ = cursor.Close(ctx) }()
 
-	var docsToDelete []bson.M
-	if err = cursor.All(ctx, &docsToDelete); err != nil {
-		return 0, err
+	// Process documents in batches to avoid potential memory issues
+	const batchSize = 10000
+	var totalDeleted int64 = 0
+
+	for {
+		batch := make([]string, 0, batchSize)
+		batchCount := 0
+
+		for cursor.Next(ctx) && batchCount < batchSize {
+			var doc struct {
+				ID string `bson:"_id"`
+			}
+			if err = cursor.Decode(&doc); err != nil {
+				return totalDeleted, err
+			}
+			batch = append(batch, doc.ID)
+			batchCount++
+		}
+
+		if len(batch) == 0 {
+			break
+		}
+
+		// Delete the batch
+		result, err := p.Collection.DeleteMany(ctx, bson.M{"_id": bson.M{"$in": batch}})
+		if err != nil {
+			return totalDeleted, err
+		}
+
+		totalDeleted += result.DeletedCount
+
+		if cursor.Err() != nil {
+			return totalDeleted, cursor.Err()
+		}
+
+		// If we didn't fill the batch, we're done
+		if batchCount < batchSize {
+			break
+		}
 	}
 
-	if len(docsToDelete) == 0 {
-		return 0, nil
-	}
-
-	ids := make([]any, len(docsToDelete))
-	for i := range docsToDelete {
-		ids[i] = docsToDelete[i]["_id"]
-	}
-
-	result, err := p.Collection.DeleteMany(ctx, bson.M{"_id": bson.M{"$in": ids}})
-	if err != nil {
-		return 0, err
-	}
-
-	return result.DeletedCount, nil
+	return totalDeleted, nil
 }
 
 func (p PostCollection) GetMaxSequence(ctx context.Context) (*int64, error) {