High concurrency Architecture - distributed search engine (ES)

Distributed search engine —— ElasticSearch, Referred to as ES.

ElasticSearch The design idea is distributed search engine , The bottom is based on lucene Of . The core idea is to start multiple... On multiple machines ES Process instance , Formed a ES colony .

ES In which data is stored The basic unit is the index , For example, you need to be in ES Store some order data in , You should be in ES Create an index in  order_idx , All order data will be written into this index , An index is pretty much the same mysql A watch in .

index -> type -> mapping -> document -> field

In order to make a more direct introduction , Make an analogy . But bear in mind that Don't equal , Analogy is only for the convenience of understanding .

index amount to mysql A watch in . and type Can't follow mysql Go inside and compare , One index There can be more than one type, Every type The fields of are almost the same , But there are some slight differences . Let's say I have a index, It's an order index, It's dedicated to order data . It's like saying you're mysql CSCEC table , Some orders are orders for physical goods , Like a dress 、 A pair of shoes ; Some orders are virtual goods orders , For example, the game point card , More credit . Most fields of the two orders are the same , But a small number of fields may differ slightly .

So it will be in order index in , Build two type, One is the physical goods order type, One is virtual goods orders type, these two items. type Most of the fields are the same , A few fields are different .

In many cases , One index There may be just one type, But it's true if it's a index There are more than one type