Elasticsearch Queries

Elasticsearch Query DSL for complex searching.

Match¶

GET /articles/_search
{
  "query": {
    "match": {
      "title": "elasticsearch tutorial"
    }
  }
}

Bool¶

{
  "query": {
    "bool": {
      "must": [{"match": {"content": "kubernetes"}}],
      "filter": [
        {"term": {"tags": "devops"}},
        {"range": {"published_at": {"gte": "2024-01-01"}}}
      ]
    }
  }
}

Aggregations¶

{
  "size": 0,
  "aggs": {
    "by_tag": {"terms": {"field": "tags", "size": 20}},
    "avg_views": {"avg": {"field": "views"}}
  }
}

Practical Query Tips¶

In production environments, always use filter context instead of must for conditions that do not need scoring (date, status, type). Filters are cached and significantly faster. A bool query combining must for relevance and filter for constraints is the most common pattern.

For paginating large result sets, avoid deep pagination (from + size > 10,000). Instead, use search_after with the point-in-time API, which is more efficient and consistent. Aggregations are a powerful tool for analytics — they allow you to calculate statistics, create histograms, and run bucket analyses directly in Elasticsearch without needing to export data to an external tool.

Query DSL = Powerful¶

Bool queries + aggregations cover most needs.