Skip to main content

Search

IronERP provides various search facilities in different contexts.

Search Types​

Using global search, you can search for any entity by Name.

Advanced search for entities under a specific model

Backend​

IronERP uses Meilisearch as the search engine. All relevant database contents are automatically indexed on start.

warning

Indexing a large database can take a long time. For larger deployments, you should use persistent storage for your Meilisearch instance(s) and configure the IronERP indexer for incremental indexing.

API​

The backend does not directly expose any search-related endpoints. The frontend should always talk directly to the search engine for optimal performance. Please refer to the Meilisearch Docs for more information.

Configuration​

This section outlines the relevant configuration options for the Search functionality. Below is a complete example of the search configuration block with notes explaining each configuration option.

{
    "Search": {
        "Backend": {
            "MeilisearchHost": "http://198.51.100.1:7700", // (1)
            "SearchKey": "abc123", // (2)
            "WriteKey": "456def"     // (3)
        },
        "Indexing": {
            "Strategy": "incremental" // (4)
        }
    }
}
  1. The Meilisearch API host
  2. The read-only search key that is passed to the frontend
  3. The read-write key used for indexing documents
  4. The indexing strategy. See below for more info.

Indexing Strategy​

IronERP supports two indexing strategies out of the box.

full (default)​

With the full indexing strategy, IronERP will drop all indexes (if they exist) on start and index all documents from the database. This is useful for development as it helps ensure you always have fresh data in the search engine. Not recommended for production deployments as this could make startup very slow with larger databases.

incremental​

With the incremental strategy, IronERP will use the timestamps and IDs of indexed documents and other heuristics to determine what data needs to be indexed and only indexes documents that aren't already present in the search engine.

For optimal performance, your Meilisearch instance needs to be configured with persistent storage.