Index Settings Configuration for DataHub Simple REST Bundle

Overview

The DataHub Simple REST bundle allows you to configure Elasticsearch/OpenSearch index settings directly via Symfony configuration. This enables you to customize analyzers, tokenizers, and other index-level settings for all indices created by the bundle, while still providing robust and sensible defaults.

Configuration Structure

Index settings are now defined under the indexing_options.index_settings key in your Symfony config (e.g., config/enterprise/data_hub.yaml):

pimcore_data_hub_simple_rest:
    indexing_options:
        index_settings:
            mapping:
                total_fields:
                    limit: 2000
            number_of_replicas: 0
            max_ngram_diff: 30
            analysis:
                analyzer:
                    path_analyzer:
                        tokenizer: path_tokenizer
                    datahub_ngram_analyzer:
                        tokenizer: datahub_ngram_tokenizer
                        filter: [lowercase]
                    datahub_whitespace_analyzer:
                        tokenizer: datahub_whitespace_tokenizer
                        filter: [lowercase]
                tokenizer:
                    path_tokenizer:
                        type: path_hierarchy
                    datahub_ngram_tokenizer:
                        type: ngram
                        min_gram: 3
                        max_gram: 25
                        token_chars: [letter, digit]
                    datahub_whitespace_tokenizer:
                        type: whitespace

You can override any of these settings in your project config. Any omitted values will fall back to the bundle's built-in defaults.

How It Works

• The bundle uses a Symfony VariableNodeDefinition for index_settings, allowing you to specify any valid Elasticsearch/OpenSearch index settings.
• The settings are injected into all index creation and update operations.

Example: Overriding a Setting

To change the number of replicas for all indices:

pimcore_data_hub_simple_rest:
    indexing_options:
        index_settings:
            number_of_replicas: 2