Speeding Up Magento 2 Search with Elasticsearch 7.x: A Practical Guide

Problem: Slow, Inaccurate Search in Magento 2 Stores

Many Magento 2 shops still rely on MySQL full‑text search. On medium‑to‑large catalogs it becomes sluggish and often misses relevant results. The default MySQL engine also lacks advanced faceting and relevance tuning that modern shoppers expect.

Thesis: Switch to Magento’s Default Elasticsearch 7.x Engine and Fine‑Tune It

Magento 2.4+ ships with Elasticsearch 7.x as the built‑in search engine. Enabling it, configuring the indexers, and tweaking relevance can cut search latency by up to 30% and dramatically improve hit‑rate. The following steps walk you through the process, with a concrete example and a clear trade‑off discussion.

1. Enable Elasticsearch in Magento 2

1.1. Verify Compatibility

Before making any changes, confirm that your Magento version supports the Elasticsearch major version you plan to use. For Magento 2.4.5, Elasticsearch 7.9 is supported; newer 7.x releases may break indexing. Check the Magento DevDocs for the exact mapping.

1.2. Configure the Connection

There are two common ways to set the connection details:

1. Admin UI – Navigate to Stores > Configuration > Catalog > Catalog Search and fill in the host, port, and cluster name.
2. env.php – Edit app/etc/env.php to add the elasticsearch array. This method is preferred for production because it keeps credentials out of the admin UI.

Example env.php snippet:

'search' => [
    'engine' => 'elasticsearch7',
    'elasticsearch7' => [
        'host' => '10.0.0.5',
        'port' => '9200',
        'index_prefix' => 'magento2',
        'cluster' => 'prod-cluster',
        'user' => 'elastic',
        'password' => 'secret',
    ],
],

After editing env.php, run bin/magento setup:upgrade to apply the changes.

1.3. Verify Connection

Use the cluster health API to ensure Magento can talk to Elasticsearch:

curl http://10.0.0.5:9200/_cluster/health?pretty

Expected output: status: "green". If you see status: "red" or connection errors, fix network or credentials first.

2. Configure Magento Indexers for Elasticsearch

2.1. Indexing Mode

Magento offers two modes: Update on Save (real‑time) and Update by Schedule (batch). For most shops, Update on Save keeps search fresh but increases write load. Use Update by Schedule during high‑traffic periods.

Set it via CLI:

bin/magento indexer:set-mode schedule search_catalog_product

Then schedule a cron job to run indexer:reindex every 15 minutes.

2.2. Dedicated Search Indexer

Magento automatically creates a catalogsearch_fulltext index. Verify its status:

bin/magento indexer:status

All indexes should report Ready. If any show Processing or Invalid, run:

bin/magento indexer:reindex catalogsearch_fulltext

3. Tune Relevance and Faceting

3.1. Custom Analyzers

Elasticsearch’s analyzers control how text is tokenized. Magento exposes a simple configuration for common analyzers:

    
        
            standard
            standard
        
        
            my_custom
            custom
            standard
            
                lowercase
                stop
            
        
    

To apply a custom analyzer to a field, edit the etc/elastic.xml in a custom module or theme. This step is advanced and usually only needed for multilingual or niche catalogs.

3.2. Field Boosting

Boosting tells Elasticsearch to weight certain fields higher. In Magento, you can adjust the search_weight for attributes via the admin:

bin/magento catalog:attribute:set-visibility --attribute-code name --visibility 4

Alternatively, edit the catalogsearch.xml to set boost values per field.

3.3. Synonyms

Synonym files allow the search to treat different terms as equivalent. Place a synonyms.txt file in var/search and reference it in the analyzer configuration. Example entry:

color, colour, hue

After adding, reindex.

4. Practical Worked Example

Assume a store with 50,000 products and a dedicated Elasticsearch cluster at 10.0.0.5:9200. The steps below bring the search up to speed.

1. Update app/etc/env.php with the connection details shown earlier.
2. Run bin/magento setup:upgrade and bin/magento setup:di:compile to apply changes.
3. Set indexing mode to schedule for catalogsearch_fulltext:

bin/magento indexer:set-mode schedule catalogsearch_fulltext

4. Configure a cron job that runs bin/magento indexer:reindex catalogsearch_fulltext every 15 minutes.
5. Verify cluster health: curl http://10.0.0.5:9200/_cluster/health?pretty → status: "green".
6. Check indexer status: bin/magento indexer:status → all Ready.
7. Test search latency via the browser’s Network panel: before, a product search took ~1.2 s; after enabling Elasticsearch, it drops to ~0.8 s.

5. Trade‑Offs and Limitations

• Version Compatibility: Using Elasticsearch 8.x with Magento 2.4.x will break indexing. Stick to the supported 7.x range.
• Resource Overhead: Elasticsearch requires dedicated CPU, RAM, and disk I/O. Monitor vm.max_map_count and max_file_descriptors per the official guide.
• Indexing Load: Update on Save keeps search fresh but can spike CPU during bulk product uploads. Prefer Update by Schedule for large catalogs.
• Complexity of Custom Analyzers: Tweaking analyzers can improve relevance but may introduce bugs if not thoroughly tested.

6. Actionable Checklist

1. Confirm Magento‑Elasticsearch version compatibility.
2. Set connection details in env.php or Admin UI.
3. Run setup:upgrade and reindex.
4. Verify cluster health and indexer status.
5. Measure search latency pre‑ and post‑configuration.
6. Adjust indexing mode and analyzer settings based on performance data.
7. Document the configuration for future maintenance.

By following these steps, you can transform Magento’s search from a bottleneck into a competitive advantage, delivering faster, more relevant results to your customers.