Elasticsearch
ELI5 - The Vibe Check
Elasticsearch is what you reach for when your database's built-in search is basically does-this-text-contain-that-exact-substring, and you actually need something that understands typos, relevance ranking, and searching across millions of documents in milliseconds. It's the engine quietly powering the search bar behind a huge chunk of the internet's type-something-and-get-good-results-instantly experiences.
Real Talk
Elasticsearch is a distributed, JSON-based search and analytics engine built on Apache Lucene, designed for full-text search, log aggregation, and near-real-time analytics at scale. It indexes documents into shards distributed across a cluster, supports fuzzy matching, relevance scoring, and complex aggregations through its query DSL, and commonly forms the E in the ELK stack alongside Logstash and Kibana for observability pipelines.
Show Me The Code
GET /products/_search
{
"query": {
"match": { "name": "wireles headphones" }
}
}
When You'll Hear This
"Postgres full-text search wasn't cutting it, we moved product search to Elasticsearch." / "Check Kibana, the Elasticsearch cluster is showing yellow health again."
Related Terms
Algolia
Algolia is the luxury hotel of search engines.
Full-Text Search
Full-text search lets you search through text like Google does, not just exact matches.
Fuzzy Search
Fuzzy search finds results even when the search term is misspelled or slightly off. Search for 'jonh' and it still finds 'John.
Grafana Loki
Grafana Loki is log aggregation that doesn't cost a fortune.
Meilisearch
You want search-as-you-type on your app. Elasticsearch needs a PhD and a dedicated ops team. Meilisearch needs about 15 minutes and one Docker command.