Skip to content

Latest commit

 

History

History
409 lines (334 loc) · 18.8 KB

usage.asciidoc

File metadata and controls

409 lines (334 loc) · 18.8 KB

Getting started

This section describes how to get started with the low-level REST client from getting the artifact to using it in an application.

Javadoc

The javadoc for the low level REST client can be found at {rest-client-javadoc}/index.html.

Maven Repository

The low-level Java REST client is hosted on Maven Central. The minimum Java version required is 1.7.

The low-level REST client is subject to the same release cycle as Elasticsearch. Replace the version with the desired client version, first released with 5.0.0-alpha4. There is no relation between the client version and the Elasticsearch version that the client can communicate with. The low-level REST client is compatible with all Elasticsearch versions.

If you are looking for a SNAPSHOT version, the Elastic Maven Snapshot repository is available at https://snapshots.elastic.co/maven/.

Maven configuration

Here is how you can configure the dependency using maven as a dependency manager. Add the following to your pom.xml file:

    org.elasticsearch.client
    elasticsearch-rest-client
    {version}

Gradle configuration

Here is how you can configure the dependency using gradle as a dependency manager. Add the following to your build.gradle file:

dependencies {
    compile 'org.elasticsearch.client:elasticsearch-rest-client:{version}'
}

Dependencies

The low-level Java REST client internally uses the Apache Http Async Client to send http requests. It depends on the following artifacts, namely the async http client and its own transitive dependencies:

  • org.apache.httpcomponents:httpasyncclient

  • org.apache.httpcomponents:httpcore-nio

  • org.apache.httpcomponents:httpclient

  • org.apache.httpcomponents:httpcore

  • commons-codec:commons-codec

  • commons-logging:commons-logging

Shading

In order to avoid version conflicts, the dependencies can be shaded and packaged within the client in a single JAR file (sometimes called an "uber JAR" or "fat JAR"). Shading a dependency consists of taking its content (resources files and Java class files) and renaming some of its packages before putting them in the same JAR file as the low-level Java REST client. Shading a JAR can be accomplished by 3rd-party plugins for Gradle and Maven.

Be advised that shading a JAR also has implications. Shading the Commons Logging layer, for instance, means that 3rd-party logging backends need to be shaded as well.

Maven configuration

Here is a configuration using the Maven Shade plugin. Add the following to your pom.xml file:

    
        
            org.apache.maven.plugins
            maven-shade-plugin
            3.1.0
            
                
                    package
                    shade
                    
                        
                            
                                org.apache.http
                                hidden.org.apache.http
                            
                            
                                org.apache.logging
                                hidden.org.apache.logging
                            
                            
                                org.apache.commons.codec
                                hidden.org.apache.commons.codec
                            
                            
                                org.apache.commons.logging
                                hidden.org.apache.commons.logging
                            
                        
                    
                
            
        
    

Gradle configuration

Here is a configuration using the Gradle ShadowJar plugin. Add the following to your build.gradle file:

shadowJar {
    relocate 'org.apache.http', 'hidden.org.apache.http'
    relocate 'org.apache.logging', 'hidden.org.apache.logging'
    relocate 'org.apache.commons.codec', 'hidden.org.apache.commons.codec'
    relocate 'org.apache.commons.logging', 'hidden.org.apache.commons.logging'
}

Initialization

A RestClient instance can be built through the corresponding RestClientBuilder class, created via RestClient#builder(HttpHost…​) static method. The only required argument is one or more hosts that the client will communicate with, provided as instances of HttpHost as follows:

include-tagged::{doc-tests}/RestClientDocumentation.java[rest-client-init]

The RestClient class is thread-safe and ideally has the same lifecycle as the application that uses it. It is important that it gets closed when no longer needed so that all the resources used by it get properly released, as well as the underlying http client instance and its threads:

include-tagged::{doc-tests}/RestClientDocumentation.java[rest-client-close]

RestClientBuilder also allows to optionally set the following configuration parameters while building the RestClient instance:

include-tagged::{doc-tests}/RestClientDocumentation.java[rest-client-init-default-headers]
  1. Set the default headers that need to be sent with each request, to prevent having to specify them with each single request

include-tagged::{doc-tests}/RestClientDocumentation.java[rest-client-init-max-retry-timeout]
  1. Set the timeout that should be honoured in case multiple attempts are made for the same request. The default value is 30 seconds, same as the default socket timeout. In case the socket timeout is customized, the maximum retry timeout should be adjusted accordingly

include-tagged::{doc-tests}/RestClientDocumentation.java[rest-client-init-failure-listener]
  1. Set a listener that gets notified every time a node fails, in case actions need to be taken. Used internally when sniffing on failure is enabled.

include-tagged::{doc-tests}/RestClientDocumentation.java[rest-client-init-node-selector]
  1. Set the node selector to be used to filter the nodes the client will send requests to among the ones that are set to the client itself. This is useful for instance to prevent sending requests to dedicated master nodes when sniffing is enabled. By default the client sends requests to every configured node.

include-tagged::{doc-tests}/RestClientDocumentation.java[rest-client-init-request-config-callback]
  1. Set a callback that allows to modify the default request configuration (e.g. request timeouts, authentication, or anything that the org.apache.http.client.config.RequestConfig.Builder allows to set)

include-tagged::{doc-tests}/RestClientDocumentation.java[rest-client-init-client-config-callback]
  1. Set a callback that allows to modify the http client configuration (e.g. encrypted communication over ssl, or anything that the org.apache.http.impl.nio.client.HttpAsyncClientBuilder allows to set)

Performing requests

Once the RestClient has been created, requests can be sent by calling either performRequest or performRequestAsync. performRequest is synchronous and will block the calling thread and return the Response when the request is successful or throw an exception if it fails. performRequestAsync is asynchronous and accepts a ResponseListener argument that it calls with a Response when the request is successful or with an Exception if it4 fails.

This is synchronous:

include-tagged::{doc-tests}/RestClientDocumentation.java[rest-client-sync]
  1. The HTTP method (GET, POST, HEAD, etc)

  2. The endpoint on the server

And this is asynchronous:

include-tagged::{doc-tests}/RestClientDocumentation.java[rest-client-async]
  1. The HTTP method (GET, POST, HEAD, etc)

  2. The endpoint on the server

  3. Handle the response

  4. Handle the failure

You can add request parameters to the request object:

include-tagged::{doc-tests}/RestClientDocumentation.java[rest-client-parameters]

You can set the body of the request to any HttpEntity:

include-tagged::{doc-tests}/RestClientDocumentation.java[rest-client-body]
Important
The ContentType specified for the HttpEntity is important because it will be used to set the Content-Type header so that Elasticsearch can properly parse the content.

You can also set it to a String which will default to a ContentType of application/json.

include-tagged::{doc-tests}/RestClientDocumentation.java[rest-client-body-shorter]

RequestOptions

The RequestOptions class holds parts of the request that should be shared between many requests in the same application. You can make a singleton instance and share it between all requests:

include-tagged::{doc-tests}/RestClientDocumentation.java[rest-client-options-singleton]
  1. Add any headers needed by all requests.

  2. Customize the response consumer.

addHeader is for headers that are required for authorization or to work with a proxy in front of Elasticsearch. There is no need to set the Content-Type header because the client will automatically set that from the HttpEntity attached to the request.

You can set the NodeSelector which controls which nodes will receive requests. NodeSelector.NOT_MASTER_ONLY is a good choice.

You can also customize the response consumer used to buffer the asynchronous responses. The default consumer will buffer up to 100MB of response on the JVM heap. If the response is larger then the request will fail. You could, for example, lower the maximum size which might be useful if you are running in a heap constrained environment like the example above.

Once you’ve created the singleton you can use it when making requests:

include-tagged::{doc-tests}/RestClientDocumentation.java[rest-client-options-set-singleton]

You can also customize these options on a per request basis. For example, this adds an extra header:

include-tagged::{doc-tests}/RestClientDocumentation.java[rest-client-options-customize-header]

Multiple parallel asynchronous actions

The client is quite happy to execute many actions in parallel. The following example indexes many documents in parallel. In a real world scenario you’d probably want to use the _bulk API instead, but the example is illustative.

include-tagged::{doc-tests}/RestClientDocumentation.java[rest-client-async-example]
  1. Process the returned response

  2. Handle the returned exception, due to communication error or a response with status code that indicates an error

Reading responses

The Response object, either returned by the synchronous performRequest methods or received as an argument in ResponseListener#onSuccess(Response), wraps the response object returned by the http client and exposes some additional information.

include-tagged::{doc-tests}/RestClientDocumentation.java[rest-client-response2]
  1. Information about the performed request

  2. The host that returned the response

  3. The response status line, from which you can for instance retrieve the status code

  4. The response headers, which can also be retrieved by name though getHeader(String)

  5. The response body enclosed in an org.apache.http.HttpEntity object

When performing a request, an exception is thrown (or received as an argument in ResponseListener#onFailure(Exception) in the following scenarios:

IOException

communication problem (e.g. SocketTimeoutException)

ResponseException

a response was returned, but its status code indicated an error (not 2xx). A ResponseException originates from a valid http response, hence it exposes its corresponding Response object which gives access to the returned response.

Note
A ResponseException is not thrown for HEAD requests that return a 404 status code because it is an expected HEAD response that simply denotes that the resource is not found. All other HTTP methods (e.g., GET) throw a ResponseException for 404 responses unless the ignore parameter contains 404. ignore is a special client parameter that doesn’t get sent to Elasticsearch and contains a comma separated list of error status codes. It allows to control whether some error status code should be treated as an expected response rather than as an exception. This is useful for instance with the get api as it can return 404 when the document is missing, in which case the response body will not contain an error but rather the usual get api response, just without the document as it was not found.

Note that the low-level client doesn’t expose any helper for json marshalling and un-marshalling. Users are free to use the library that they prefer for that purpose.

The underlying Apache Async Http Client ships with different org.apache.http.HttpEntity implementations that allow to provide the request body in different formats (stream, byte array, string etc.). As for reading the response body, the HttpEntity#getContent method comes handy which returns an InputStream reading from the previously buffered response body. As an alternative, it is possible to provide a custom org.apache.http.nio.protocol.HttpAsyncResponseConsumer that controls how bytes are read and buffered.

Logging

The Java REST client uses the same logging library that the Apache Async Http Client uses: Apache Commons Logging, which comes with support for a number of popular logging implementations. The java packages to enable logging for are org.elasticsearch.client for the client itself and org.elasticsearch.client.sniffer for the sniffer.

The request tracer logging can also be enabled to log every request and corresponding response in curl format. That comes handy when debugging, for instance in case a request needs to be manually executed to check whether it still yields the same response as it did. Enable trace logging for the tracer package to have such log lines printed out. Do note that this type of logging is expensive and should not be enabled at all times in production environments, but rather temporarily used only when needed.