Website ElasticSearch Configuration
The WebsiteElasticSearchBundle uses the same parameters from parameters.yml as ElasticSearchBundle (host configuration and access credentials), but instead of the same index name it adds new parameter to parameters.yml:
- website_search_engine_index_prefix - prefix of the indexes in Elasticsearch used to store website index data.
This way backend and website search data can be stored in separate indexes and may be processed separately.
The WebsiteElasticSearchBundle configuration is similar to the one for the ElasticSearchBundle, but it is stored under the different bundle alias (oro_website_search). You can also specify engine name and engine parameters in the config.yml file:
oro_website_search: engine: "%search_engine_name%" engine_parameters: 
Like in the ElasticSearchBundle, in the WebsiteElasticSearchBundle you can specify client and index configuration. The configuration structure and format is exactly the same. Here is an example of custom index configuration:
oro_website_search: engine_parameters: client: hosts: ['%search_engine_host%:%search_engine_port%'] retries: 1 index: prefix: '%website_search_engine_index_prefix%' body: settings: index: refresh_interval: '30s'
Disabled Environment Checks
Similar to ElasticSearchBundle, you may disable system requirements check and index status check to eliminate all queries to system level routes. These checks don’t affect application performance as they are used only during the application installation of full reindexation.
Here is how these flags may be configured.
oro_website_search: engine_parameters: system_requirements_check: false index_status_check: false
Similar to ElasticSearchBundle, you may enable language optimization of indexation. It allows to use specialized language analyzers for search index based on the used localizations. If no appropriate analyzer found then default whitespace analyzer will be used instead.
Here is how language optimization may be enabled.
oro_website_search: engine_parameters: language_optimization: true
To start usage of this optimization search index must be completely removed and full reindexation has to be started to fill it with data.
Similar to ElasticSearchBundle, you may switch indexation into synchronous mode using option force_refresh at engine parameters:
oro_website_search: engine_parameters: force_refresh: true
Keep in mind that synchronous indexation is slower than asynchronous because application has to wait for reindexation to finish after every reindexation request.