This article introduces general concept of Elastic search.

ElasticSearch is a distributed, open-source search and analytics engine built on Apache lucene

Common APIs

Elasticsearch exposes a RESTful HTTP API.

Document APIs

• POST /{index}/_doc index a document
• GET /{index}/_doc/{id} get a document
• DELETE /{index}/_doc/{id} delete a document
• POST /{index}/_update/{id} update a document
• POST /{index}/_bulk execute bulk operations
• GET /{index}/_count count documents in an index
• GET /{index}/_search search documents
• POST /{index}/_search search with a request body
• POST /_msearch multi-search across one or more indices
• POST /_update_by_query update matching documents by query
• POST /_delete_by_query delete matching documents by query
• GET /{index}/_termvectors retrieve term vectors for a document

Index APIs

• PUT /{index} create an index
• DELETE /{index} delete an index
• GET /{index} get index information
• HEAD /{index} check whether an index exists
• GET /{index}/_mapping get index mapping
• PUT /{index}/_mapping update index mapping
• GET /{index}/_settings get index settings
• PUT /{index}/_settings update index settings
• POST /{index}/_refresh refresh index
• POST /{index}/_flush flush index
• POST /{index}/_forcemerge force merge index segments
• POST /{index}/_open open an index
• POST /{index}/_close close an index

Cluster APIs

• GET /_cluster/health get cluster health
• GET /_cluster/stats get cluster metrics
• GET /_cluster/state get cluster state
• GET /_cluster/settings get cluster settings
• PUT /_cluster/settings update cluster settings
• GET /_nodes get node information
• GET /_nodes/stats get node statistics
• GET /_nodes/hot_threads show hot threads on nodes
• GET /_cluster/pending_tasks list pending cluster tasks

Cat APIs

• GET /_cat/health?v cluster health in a compact format
• GET /_cat/indices?v list indices
• GET /_cat/nodes?v list nodes
• GET /_cat/shards?v list shards
• GET /_cat/allocation?v show shard allocation
• GET /_cat/recovery?v show recovery status
• GET /_cat/pending_tasks?v show pending tasks
• GET /_cat/templates?v list index templates

Snapshot and restore APIs

• PUT /_snapshot/{repository}/{snapshot} create a snapshot
• GET /_snapshot/{repository}/{snapshot} get snapshot information
• DELETE /_snapshot/{repository}/{snapshot} delete a snapshot
• POST /_snapshot/{repository}/{snapshot}/_restore restore a snapshot

Reindexing and scroll APIs

• POST /_reindex copy documents from one index to another
• GET /_search/scroll retrieve the next scroll page
• DELETE /_search/scroll clear scroll contexts

Reference

API DOCS

Mapping

Schema definition specifying how document fields are stored and indexed in Elasticsearch.

search

Elasticsearch search requests use JSON bodies to specify query logic, pagination, sorting, aggregations, and result shaping. Common fields in the search JSON include:

• query - selects documents.
  • bool - combine clauses: must, should, must_not, filter
  • match - full-text match on analyzed fields
  • term - exact value match
  • range - numeric or date intervals
  • nested - queries inside nested object fields
  • exists - field presence
• aggs / aggregations - summarize matching documents.
  • bucket aggs: terms, range, date_histogram, histogram, filters
  • metric aggs: avg, sum, min, max, stats, cardinality, value_count
• size - hits per page
• from - result offset for pagination
• sort - result ordering by field or score
• _source - which document fields to return
• fields - stored/scripted field values to fetch
• highlight - return matched snippets
• post_filter - filter search hits after aggregations
• timeout - time limit for the search
• track_total_hits - accurate hit count control
• explain - return scoring explanation for each hit
• track_scores - keep scores when sorting/paging
• profile - return query execution timing
• search_after - deep pagination cursor using sort values
• collapse - collapse results by a field value
• suggest - autocomplete or correction suggestions
• script_fields - compute extra fields per hit with scripts
• rescore - rerank top hits with a second query

These fields can be combined to express complex search behavior.