Compatibility Guide

更新时间:
复制 MD 格式

This document describes the compatibility of the Lindorm search engine with open source Elasticsearch and OpenSearch clients, ecosystem components, and plugins. Use it to quickly identify supported version combinations.

Client Compatibility

The Lindorm search engine supports both Elasticsearch and OpenSearch clients. Follow these recommendations:

Java Clients

Client Type

Recommended Version

Version Range

Compatibility

ElasticSearch High Level REST Client

7.10.x

7.x (7.0.x–7.17.x)

Compatible

ElasticSearch Low Level REST Client

OpenSearch Java Client

2.7

2.x (2.0.x–2.26.x)

Compatible

For detailed Java client integration instructions and code examples, see Connect to the search engine using the Java API.

Python Clients

Client Type

Recommended Version

Version Range

Compatibility

ElasticSearch Python Client

7.10.x

7.0.0–7.13.4

Compatible

7.13.4 and later up to 7.17.x

Submit a ticket to contact technical support for whitelist configuration

OpenSearch Python Client

2.7

All 2.x versions

Compatible

Ecosystem Component Compatibility

Kibana

The Lindorm search engine recommends using the built-in OpenSearch Dashboards, which provide basic features such as visualization and the development console. If you must connect a self-deployed Kibana instance to the Lindorm search engine, refer to the following compatibility information:

Kibana Version Type

Compatibility

Recommended Version

Kibana-oss

Compatible

7.10.2

Non-oss versions

Not compatible

Version checks prevent direct connection to the Lindorm search engine

Logstash and Filebeat

Logstash and Filebeat follow the same compatibility rules, described below.

Version Range

Compatibility

Recommended Version

7.x versions up to and including 7.13.4

Compatible

7.10.2

7.x versions above 7.13.4

Submit a ticket to contact technical support for whitelist configuration

-

8.x

-

Server-Side Configuration Recommendations

Enable automatic index creation to allow Filebeat or Logstash to automatically create indexes when writing data to the Lindorm search engine. Also, configure index templates so newly created indexes inherit consistent settings, such as mapping definitions and shard counts.

Run the following request using the OpenSearch Dashboards development console or a curl command:

PUT _cluster/settings
{
  "persistent": {
    "action.auto_create_index": "true"
  }
}

After configuration, run this request to confirm the setting is active:

GET _cluster/settings

If the value of persistent.action.auto_create_index in the response is "true", the configuration succeeded.

Client Configuration Recommendations

Disable automatic template creation and index lifecycle management (ILM) in your Filebeat or Logstash task configuration.

Filebeat configuration example:

setup.template.enabled: false
setup.ilm.enabled: false

Logstash configuration example:

output {
  elasticsearch {
    hosts => ["http://<Lindorm search engine endpoint>"]
    ...
    manage_template => false
    ilm_enabled => false
  }
}

Plugin Compatibility

Custom Plugin Requirements

The Lindorm search engine supports custom plugins that meet these conditions:

  • Plugin class inheritance requirement: The plugin class must inherit from org.opensearch.plugins.Plugin. Plugins built on the Elasticsearch plugin framework are not supported.

  • Version consistency requirement: The opensearch.version field in the plugin descriptor file (plugin-descriptor.properties) must be set to 2.7.0.

Built-In Tokenizer Plugins

The Lindorm search engine includes the following tokenizer plugins. You can use them without additional installation:

Plugin Name

Description

search-analisys-pinyin

Pinyin tokenizer

analysis-ik

IK tokenizer

analysis-icu

ICU tokenizer

analysis-smartcn

SmartCN tokenizer

jieba

jieba tokenizer

Note

If you encounter any compatibility issues or have questions, contact Lindorm technical support (DingTalk ID: s0s3eg3).

Vector Retrieval Connection Guide

For vector retrieval code examples, see the following links: