Truncate Collection

Truncating a collection removes all entities while preserving the collection's schema, constraints, and indexes. It is more efficient than deleting entities because it hides all entities flushed before the current timestamp from searches and queries and drops them in the background.

📘Notes

This feature applies only to managed collections.

Overview

Collection truncation is a high-performance operation that removes all entities from a collection while fully preserving its structural definition, including schema, constraints, and indexes. This keeps the collection immediately ready for new data ingestion without requiring reconfiguration or index rebuilding.

Unlike conventional deletion methods that process records individually and generate extensive transaction logs, truncation operates through an optimized two-step mechanism:

1. Immediate logical removal

   All entities inserted or deleted before the truncation timestamp are immediately flushed and hidden from searches and queries, making them effectively invisible to subsequent operations.

2. Efficient physical cleanup

   The system garbage-collects all affected data segments in the background, eliminating the overhead of per-entity deletion processing.

Truncation is ideal for use cases that require rapid, complete dataset resets, such as test environment refreshes, pipeline stage cleanup, or periodic data lifecycle management, where performance and resource efficiency are critical.

Example

The following code examples assume that you already have a collection named my_collection.

Python

python

from pymilvus import MilvusClient, DataType

client = MilvusClient(
    uri="YOUR_CLUSTER_ENDPOINT",
    token="YOUR_CLUSTER_TOKEN"
)

client.truncate_collection(
    collection_name="my_collection"
)

Java

java

// java
import io.milvus.v2.client.ConnectConfig;
import io.milvus.v2.client.MilvusClientV2;
import io.milvus.v2.service.collection.request.TruncateCollectionReq;

public class TruncateExample {
  public static void main(String[] args) {
      ConnectConfig connectConfig = ConnectConfig.builder()
              .uri("YOUR_CLUSTER_ENDPOINT")
              .token("YOUR_CLUSTER_TOKEN")
              .build();
      MilvusClientV2 client = new MilvusClientV2(connectConfig);

      // Truncate collection
      TruncateCollectionReq req = TruncateCollectionReq.builder()
              .collectionName("my_collection")
              .build();
      client.truncateCollection(req);

      System.out.println("collection truncated successfully");
      client.close();
  }
}

Go

go

// go
package main

import (
    "context"
    "fmt"
    "log"

    "github.com/milvus-io/milvus/client/v2/milvusclient"
)

func main() {
    ctx := context.Background()

    client, err := milvusclient.New(ctx, &milvusclient.ClientConfig{
            Address: "YOUR_CLUSTER_ENDPOINT",
            APIKey: "YOUR_CLUSTER_TOKEN"
    })
    if err != nil {
            log.Fatal("failed to connect:", err)
    }
    defer client.Close(ctx)

    err = client.TruncateCollection(ctx, milvusclient.NewTruncateCollectionOption("my_collection"))
    if err != nil {
            log.Fatal("failed to truncate:", err)
    }

    fmt.Println("collection truncated successfully")
}

NodeJS

javascript

const milvusClient = new MilvusClient({ 
    address: 'YOUR_CLUSTER_ENDPOINT', 
    token: 'YOUR_CLUSTER_TOKEN'
});

const res = await milvusClient.truncateCollection({
    collection_name: my_collection,
 });

cURL

bash

export CLUSTER_ENDPOINT="YOUR_CLUSTER_ENDPOINT"
export TOKEN="YOUR_CLUSTER_TOKEN"

# restful
curl -X POST "${CLUSTER_ENDPOINT}/v2/vectordb/collections/truncate" \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer ${TOKEN}" \
  -H "Request-Timeout: 10" \
  -d '{
    "dbName": "default",
    "collectionName": "my_collection"
  }'

C++

c++

// C++