ElasticSearch For SQL People Cheat Sheet (DRAFT) by Louis Cribbins

SQL	Elasti­cSearch
Database	Index
Table	Type
Row	Document
Column	Field

Index	Collection of one or more shards
Shard	Apache Lucene Instance
Cluster	One or more nodes
Nodes	Server instances
Routing	A way to control what shard a document goes to (used for perfor­mance)
Segment	Internal lucene data block

Shard(s)	One Primary, N Replicas
Index	Made of one or more shards
Cluster	Made of one or more nodes
Nodes	A single server in a cluster
Shard	Logical partition of an index across the nodes in a cluster
Replica	Number of copies of the primary shards in a cluster
Total shards in cluster	= primary shards * (replica count + 1)
Shards per node	= Total shards in cluster / # of nodes

Elasti­cSearch	Apache Lucene
Node management	Searching
Shard management	Indexing
Health monitoring	Stemming
Parall­elizing Work	Serialized Workload

Sharding	Children reside in Parent's shard
Parent­-Child	Cheap to update, slow to query
Nested	Cheap to query, slow to update

PUT server­/in­dex­/ty­pe/ID JSON	Creates doc from JSON using ID
POST server­/in­dex­/type JSON	Creates doc from JSON and generates ID
POST server­/in­dex­/ty­pe/­ID/­_update JSON	Updates part of the document matching on ID
GET server­/in­dex­/ty­pe/ID	Retrieves doc matching ID
DELETE server­/in­dex­/ty­pe/ID	Deletes doc matching ID

Forward Index	Maps a document to a set of words
Inverted Index	Maps a word to a set of documents
Stemming	Converting a word to it's root, ie raining -> rain
Synonyms	Converts a word to an approp­riate equiva­lent, ie Sunday -> weekend