Before launching queries, ensure your server environment permits necessary cross-origin requests and establish a management interface.

1. CORS Configuration: In the config/elasticsearch.yml file of your installation directory, enable HTTP Cross-Origin Resource Sharing to allow frontend access:

# Enable HTTP CORS support
http.cors.enabled: true
# Allow all origins (use '*' cautiously)
http.cors.allow-origin: "*"

2. Installation Manager Tool: To visualize clusters and documents, deploy the elasticsearch-head plugin via NPM.

# Install global CLI tools
npm i -g grunt-cli

# Navigate to project root and run installer
npm i .

# Launch the server interface
grunt server

Core Indexing Operations

The following steps demonstrate managing an index named articles_db, including schema definition and document manipulation.

• Create Index: Define shard count, replica settings, and field mappings.

PUT localhost:9200/articles_db
Content-Type: application/json
{
  "settings": {
    "index": {
      "number_of_shards": 3,
      "number_of_replicas": 1
    }
  },
  "mappings": {
    "properties": {
      "doc_uid": {
        "type": "keyword"
      },
      "headline": {
        "type": "text",
        "store": true,
        "analyzer": "standard"
      },
      "main_body": {
        "type": "text",
        "store": true,
        "analyzer": "standard"
      }
    }
  }
}

• Delete Index: Permanently remove the index.

DELETE localhost:9200/articles_db

• Index Document: Create or update a specific document by ID.

POST localhost:9200/articles_db/_doc/101
Content-Type: application/json
{
  "doc_uid": 101,
  "headline": "New Document Entry",
  "main_body": "This represents sample content for testing purposes"
}

• Delete Document: Remove a specific entry.

DELETE localhost:9200/articles_db/_doc/101

• Retrieve Document: Fetch details for a single record.

GET localhost:9200/articles_db/_doc/101

• Keyword Search (Term Match): Exact matching on specific tokens.

POST localhost:9200/articles_db/_search
Content-Type: application/json
{
  "query": {
    "term": {
      "main_body": "sample"
    }
  }
}

• Phrase/Query Search: Search across multiple fields with operator logic.

POST localhost:9200/articles_db/_search
Content-Type: application/json
{
  "query": {
    "query_string": {
      "default_field": "headline",
      "query": "entry testing"
    }
  }
}

• Analyze API Test: Inspect how a text segment is tokenized.

POST localhost:9200/_analyze
Content-Type: application/json
{
  "analyzer": "standard",
  "text": "This is a simple test"
}

Standard analyzers often split Chinese text inefficiently. Integrate the analysis-ik plugin to optimize search relevance for Chinese characters.

1. Download Plugin: Obtain the latest release archive from the official GitHub repository.

https://github.com/medcl/elasticsearch-analysis-ik/archive/master.zip

2. Compilation: Unzip the package and navigate to the directory to build the artifacts.

mvn package

3. Deployment: Move the compiled zip located in target/releases/ to the plugin folder (plugins/) of your Elasticsearch installation. Extract it and name the folder ik.

4. Version Verification: Open plugin-descriptor.properties inside the new plugin directory to confirm compatibility with your running Elasticsearch version number.

5. Verification: Restart the service and validate splitting granularity. Options include ik_smart (coarse) and ik_max_word (fine-grained).

POST localhost:9200/_analyze
Content-Type: application/json
{
  "analyzer": "ik_smart",
  "text": "这是一段测试文本"
}

To create a resilient cluster, multiple nodes must communicate securely. Ensure storage directories are cleaned before starting a fresh topology.

Configure the elasticsearch.yml file for each participating node to define unique identifiers and network interfaces.

# Unique identifier for the cluster group
cluster.name: my-elasticsearch-cluster

# Distinct name for this specific node instance
node.name: node-alpha

# Bind address (Use local IP or wildcard in production)
network.host: 0.0.0.0

# HTTP transport port (Must be unique per host)
http.port: 9205

# Inter-node communication port (Must be unique per host)
transport.tcp.port: 9305

# Discovery hosts for initial seed list
discovery.zen.ping.unicast.hosts: ["127.0.0.1:9305", "127.0.0.1:9306", "127.0.0.1:9307"]

# Designate the initial master node for bootstrapping
cluster.initial_master_nodes: ["127.0.0.1:9305"]

Repeat this configuration on other nodes, altering only node.name, http.port, transport.tcp.port, and the relevant master list entry to prevent conflicts.