Description: Elastic Search is a highly extensible open source full-text Search engine that searches almost in real time. Using ES as a search engine can provide a solution to the needs of complex search functions.
NRT (near real time) : There is only a short delay from adding data to being searchable.
Cluster: Multiple ES servers can be used as a cluster and can be searched on any node.
Node: A single ES server in a cluster that stores data and participates in the indexing and search functions of the cluster.
Index: a collection of documents with similar features. A database similar to Mysql. A single cluster can have any number of indexes.
Types: Earlier versions of ES allowed multiple types for an index, but later versions will not, and types will be removed in the future.
Documents: Documents are basic units of information that can be indexed.
Shard:
Indexes may store large amounts of data that may exceed the hardware limits of a single node or may be too slow to satisfy search requests from a single node on its own. Es solves this problem by sharding, and when creating an index, you only need to define the number of shards needed. Each shard is itself a fully functional and independent index that can be hosted on any node in the cluster.
Sharding helps with horizontal splitting to reduce capacity, while allowing parallel operations across sharding to improve throughput.
The mechanism for how sharding is assigned and how documents are aggregated into search results is transparent to the user.
Copy:
To avoid network and single point of failure, index shards are allowed to be copied into multiple replica shards for failover.
Replicas can improve system availability as well as throughput
By default, each index in ES is assigned a master shard and a copy.
Each ElasticSearch shard is a Lucene index.
Elasticsearch7.1 英 文 版 – chapter 1 – getting started