Elasticsearch Optimizing Elasticsearch Bulk Indexing for High Performance

By Opster Team

Updated: Jan 28, 2024

| 2 min read


Elasticsearch is a widely-used search and analytics engine that provides powerful capabilities for indexing and searching large volumes of data. One of the most common operations in Elasticsearch is indexing documents, which can be done individually or in bulk.

In this article, we will focus on optimizing the bulk indexing process for high performance. If you want to learn about Elasticsearch document, check out this guide. You should also take a look at this guide, which contains a detailed explanation on Elasticsearch bulk.  

Bulk indexing is a crucial operation when dealing with large datasets, as it allows you to add multiple documents to an index in a single request. This can significantly improve indexing performance and reduce the overhead associated with individual indexing requests. Here, we will discuss some best practices and techniques for optimizing bulk indexing in Elasticsearch.

1. Choose the Right Bulk Size

The first step in optimizing bulk indexing is to choose the right bulk size. The bulk size refers to the number of documents that are indexed in a single request. A larger bulk size can lead to better throughput, but it can also consume more memory and increase the risk of request timeouts.

To find the optimal bulk size, you can start with a small number (e.g., 1000 documents) and gradually increase it while monitoring the indexing performance. Keep in mind that the optimal bulk size may vary depending on your hardware, network, and Elasticsearch configuration.

2. Use the Bulk API

Elasticsearch provides a dedicated Bulk API for performing bulk indexing operations. The Bulk API allows you to send multiple index, update, and delete actions in a single request, which can significantly improve performance compared to individual requests.

To use the Bulk API, you need to create a bulk request that contains the actions and documents you want to index. Here’s an example of a bulk request in JSON format:

{ "index" : { "_index" : "my_index", "_id" : "1" } }
{ "field1" : "value1", "field2" : "value2" }
{ "index" : { "_index" : "my_index", "_id" : "2" } }
{ "field1" : "value3", "field2" : "value4" }

You can then send the bulk request to Elasticsearch using the following command:

curl -X POST "localhost:9200/_bulk" -H "Content-Type: application/json" --data-binary "@bulk_request.json"

3. Use Multiple Threads or Processes

To further improve bulk indexing performance, you can use multiple threads or processes to send bulk requests concurrently. This can help you utilize the full capacity of your Elasticsearch cluster and reduce the time it takes to index large datasets.

When using multiple threads or processes, make sure to monitor the performance and resource usage of your Elasticsearch cluster. If you notice high CPU or memory usage, you may need to reduce the number of concurrent requests or adjust your Elasticsearch configuration.

4. Disable Refresh and Replicas

During the bulk indexing process, Elasticsearch continuously refreshes the index and creates replicas of the indexed documents. While these operations are essential for ensuring data consistency and availability, they can also slow down the indexing process.

To speed up bulk indexing, you can temporarily disable the refresh interval and set the number of replicas to 0. This can be done using the following index settings:

PUT /my_index/_settings
  "index" : {
    "refresh_interval" : "-1",
    "number_of_replicas" : 0

After the bulk indexing is complete, you can re-enable the refresh interval and set the number of replicas to their original values.

5. Optimize Elasticsearch Configuration

Finally, you can optimize your Elasticsearch configuration to improve bulk indexing performance. Some of the key settings to consider include:

  • `indices.memory.index_buffer_size`: This setting determines the amount of memory allocated for indexing operations. Increasing the index buffer size can improve indexing performance but may also consume more memory.
  • `index.translog.durability`: This setting determines when the transaction log is flushed to disk. Setting the durability to “async” can improve indexing performance but may also increase the risk of data loss in case of a crash.


In conclusion, optimizing Elasticsearch bulk indexing requires careful consideration of various factors, including bulk size, API usage, concurrency, index settings, and configuration. By following the best practices and techniques discussed in this article, you can significantly improve the performance of your bulk indexing operations and ensure efficient handling of large datasets in Elasticsearch.