Specific methods of using java to operate elasticsearch

System environment: CentOS 7.2 under vm12

Current installed version: elasticsearch-2.4 0.tar. gz

Java operation es cluster step 1: configure cluster object information; 2: Create client; 3: View cluster information

1: Cluster name

The default cluster name is elasticsearch. If the cluster name is inconsistent with the specified, an error will be reported when using node resources.

2: Sniffing function

Via client transport. Sniff starts the sniffing function. In this way, you only need to specify a node in the cluster (not necessarily the primary node), and then load other nodes in the cluster. In this way, as long as the program keeps running, you can still connect to other nodes even if this node goes down.

3: Query type SearchType QUERY_ THEN_ FETCH

There are four types of ES query

QUERY_ AND_ FETCH:

The main node distributes the query request to all segments. Each segment scores and sorts according to its own query rule, that is, word frequency and document frequency, and then returns the results to the main node. The main node summarizes and sorts all data and then returns it to the client. In this way, it only needs to interact with ES once.

This query method has the problem of data volume and sorting. The master node will summarize the data returned by all partitions, so the data volume will be relatively large. Second, the rules on each partition may be inconsistent.

QUERY_ THEN_ FETCH:

The master node distributes the request to all partitions. After each partition is scored and sorted, the ID and score of the data are returned to the master node. After receiving it, the master node performs summary sorting, and then reads the corresponding data to the corresponding node according to the sorted ID, and then returns it to the client. In this way, it needs to interact with ES twice.

This method solves the problem of data volume, but the sorting problem still exists, and it is the default query method of ES

DEF_ QUERY_ AND_ Fetch: and DFS_ QUERY_ THEN_ FETCH:

Unify the rules of each partition for scoring. Sorting problem solved, but DFS_ QUERY_ AND_ Fetch still has data volume problem, DFS_ QUERY_ THEN_ Both kinds of fetch Oh, darling, you solve the problem, but the efficiency is the worst.

1. Get the client in two ways

PS: the two methods given on the official website can't be used. They need to be used together. It wastes me an afternoon

Meaning of other parameters:

code:

Tes2 Code:

The above is the whole content of this article. I hope it will be helpful to your study, and I hope you can support programming tips.