Elasticsearch Optimizing Schema Design in Elasticsearch: Techniques & Best Practices

By Opster Team

Updated: Jul 23, 2023

| 2 min read


Designing an efficient schema in Elasticsearch is crucial for achieving optimal performance and scalability. A well-designed schema can significantly improve query performance, reduce resource consumption, and minimize storage requirements. In this article, we will discuss some advanced techniques and best practices for optimizing schema design in Elasticsearch.

1. Choose the Right Data Types

Selecting the appropriate data types for your fields is essential for efficient indexing and querying. Elasticsearch supports various data types, such as text, keyword, date, integer, long, float, double, and more. Choose the data type that best fits your data and query requirements.

For example, if you have a field that contains a short, non-analyzed string, such as a product ID or an email address, use the keyword data type. This will enable efficient exact-match and aggregation queries. On the other hand, if you have a field that contains a large amount of text, such as a product description or a blog post, use the text data type to enable full-text search capabilities. If you don’t need scoring or positional queries (like it’s usually the case when querying log messages), you can also use the match_only_text type which is better suited to this use case than the text type.

2. Use Nested Objects and Parent-Child Relationships Wisely

Elasticsearch supports nested objects and parent-child relationships for modeling complex data structures. However, these features can impact performance and should be used judiciously.

Nested objects can be useful for modeling one-to-many relationships within a single document. However, they can increase indexing overhead and query complexity. Consider denormalizing nested objects into separate fields or using a flattened data type if the nested structure is not essential for your queries.

Parent-child relationships can be used to model separate entities with independent lifecycles. However, they can increase the complexity of indexing and querying, as well as consume additional resources for maintaining the relationships. Consider alternative approaches, such as denormalizing the data or using application-side joins, if the parent-child relationship is not critical for your use case.

3. Optimize Field Mappings

Field mappings define how Elasticsearch should index and store your data. Optimizing field mappings can significantly improve query performance and reduce storage requirements.

  • Disable indexing for fields that are not used in queries or aggregations. This will reduce indexing overhead and storage requirements.
  • Use the “index_options” parameter to control the amount of information stored in the index for text fields. For example, set “index_options” to “docs” if you only need to support match and term queries, and do not require scoring or highlighting.
  • Use the “doc_values” parameter to control whether a field should be stored in a columnar format for efficient sorting and aggregations. Disable “doc_values” for fields that are not used in sorting or aggregations to save storage space.
  • Use the “norms” parameter to control whether field-length normalization factors should be stored for text fields. Disable “norms” for fields where scoring based on field length is not important to save storage space.

4. Optimize Analyzers and Tokenizers

Analyzers and tokenizers play a crucial role in full-text search capabilities. Optimizing them can improve indexing and query performance and reduce resource consumption.

  • Choose the appropriate analyzer for your text fields based on your language and search requirements. Elasticsearch provides several built-in analyzers, such as the “standard” analyzer, “simple” analyzer, and language-specific analyzers.
  • Customize analyzers and tokenizers to better suit your specific use case. For example, you can create a custom analyzer that combines a specific tokenizer, filter, and character filter to optimize the analysis process for your data.
  • Use the “search_analyzer” and “search_quote_analyzer” parameters to define separate analyzers for indexing and searching. This can help improve query performance by using a more efficient analyzer for searching.

5. Optimize Index Settings

Index settings can have a significant impact on performance and resource consumption. Consider the following optimizations:

  • Use the “number_of_shards” and “number_of_replicas” parameters to control the number of primary and replica shards for your index. Choose the appropriate values based on your data size, query load, and hardware resources.
  • Use the “refresh_interval” parameter to control how often Elasticsearch should refresh the index to make new documents searchable. Increase the refresh interval for write-heavy workloads to reduce indexing overhead.
  • Use the “codec” parameter to control how the data is compressed.  Using “best_compression” implies a higher compression ratio of your index at the cost of a slower performance for extracting stored fields., such as _source.


In conclusion, optimizing schema design in Elasticsearch is a critical aspect of achieving high performance and scalability. By carefully selecting data types, optimizing field mappings, and fine-tuning index settings, you can significantly improve query performance, reduce resource consumption, and minimize storage requirements. Always consider your specific use case and requirements when designing your schema to ensure the best possible results.