使用go-mysql-elasticsearch同步数据后，ElasticSearch映射修改问题咨询

Hey there! Since your initial MySQL-to-Elasticsearch sync via go-mysql-elasticsearch is running smoothly, let's dive into the common pitfalls and fixes for updating the mappings on your products index (and type):

1. Check Elasticsearch Version Compatibility #

First off, keep in mind that Elasticsearch 7.x+ removed support for custom document types (like your products type). If you're running a newer ES version:

• Your index shouldn't have a custom type anymore—all documents use the default _doc type.
• Trying to modify a mapping for a custom type will throw errors. Instead, target the index directly with your PUT request:

  PUT 127.0.0.1:9200/products/_mapping
  {
    "properties": {
      "your_new_field": {
        "type": "text",
        "analyzer": "standard"
      }
    }
  }
  

If you're on ES 6.x (which still allows single types per index), double-check your request path is PUT 127.0.0.1:9200/products/products/_mapping—but note that 6.x is end-of-life, so upgrading is a good long-term fix.

2. Avoid go-mysql-elasticsearch Overwriting Your Manual Mappings #

The sync tool has built-in settings that might auto-update mappings based on MySQL schema changes, which could clash with your manual edits:

• Open your go-mysql-elasticsearch config file (usually river.toml) and check for these parameters:
  • auto_sync_schema: If set to true, the tool will resync the mapping whenever it detects MySQL schema changes. Set this to false if you want full control over ES mappings.
  • custom_index_mappings: You can define your desired mappings directly in this config section, so the tool uses your predefined structure instead of auto-generating one. This is safer than manual edits after sync starts.

3. Fix Mapping Update Syntax & Rules #

Elasticsearch has strict rules for modifying existing mappings—here's what to watch for:

• You can't change the data type of an existing field (e.g., from text to integer). If you need to do this, you'll have to:
  1. Create a new index with the correct mapping.
  2. Re-sync your data from MySQL to the new index.
  3. Swap the old index with the new one using an alias.
• When adding new fields, make sure your JSON payload is valid. For example, a valid add-field request looks like this:

  PUT 127.0.0.1:9200/products/_mapping
  {
    "properties": {
      "discount_percent": {
        "type": "float",
        "index": true
      },
      "product_tags": {
        "type": "keyword"
      }
    }
  }
  

• Always check the ES API response for specific error messages—they'll tell you exactly what's wrong (e.g., "illegal_argument_exception" if you're trying to modify a field type).

4. Check Tool Logs for Clues #

If your manual mapping changes keep getting reverted or ignored, check go-mysql-elasticsearch's logs. Look for entries about schema sync or mapping updates—this will show if the tool is overriding your edits, or if it's throwing errors when trying to apply MySQL schema changes to your custom mapping.

5. Verify Permissions #

Ensure the user account you're using to send the PUT request has the necessary permissions (like manage or index_admin) on the products index. Without these, ES will reject your mapping update request.

内容的提问来源于stack exchange，提问作者Grim Reaper