如何在DynamoDB中使用非主键字段实现类似MongoDB的查询？

Got it, let's break down how to replicate your MongoDB query logic in DynamoDB when requester_manager_id isn't the primary key. DynamoDB operates differently than MongoDB—it doesn't support ad-hoc queries on non-key attributes out of the box, so we need to use a Global Secondary Index (GSI) to make this efficient and cost-effective.

Step 1: Create a Global Secondary Index (GSI) on requester_manager_id #

First, you need to create an index that targets requester_manager_id as its partition key. This tells DynamoDB to maintain a separate sorted structure for this attribute, enabling fast, targeted queries.

You can create this index via the AWS Console, CLI, or SDK. Here's an example CLI command if your table is named Registrations:

aws dynamodb update-table \
  --table-name Registrations \
  --attribute-definitions AttributeName=requester_manager_id,AttributeType=S \
  --global-secondary-index-updates "[{\"Create\":{\"IndexName\":\"RequesterManagerIdIndex\",\"KeySchema\":[{\"AttributeName\":\"requester_manager_id\",\"KeyType\":\"HASH\"}],\"Projection\":{\"ProjectionType\":\"ALL\"},\"ProvisionedThroughput\":{\"ReadCapacityUnits\":5,\"WriteCapacityUnits\":5}}}]"

• AttributeType=S assumes requester_manager_id is a string—adjust to N if it's a number.
• ProjectionType=ALL means the index includes all attributes from the table. If you only need specific fields, use INCLUDE and list the attributes to optimize index size and performance.
• Adjust ProvisionedThroughput based on your expected query/write load.

Step 2: Query the GSI to Fetch Registration Details #

Once the index is active, you can use DynamoDB's query operation (avoid scan—it’s inefficient for large datasets) to retrieve items where requester_manager_id matches your input. Here's the Node.js/Express code equivalent to your MongoDB snippet:

const AWS = require('aws-sdk');
const dynamodb = new AWS.DynamoDB.DocumentClient();
const logger = require('./your-logger-module'); // Update to your logger path

exports.findManager = function (req, res) {
  const params = {
    TableName: 'Registrations',
    IndexName: 'RequesterManagerIdIndex', // Match the GSI name you created
    KeyConditionExpression: 'requester_manager_id = :managerId',
    ExpressionAttributeValues: {
      ':managerId': req.params.managerId
    }
  };

  dynamodb.query(params, (err, data) => {
    if (err) {
      logger.error(`Error while retrieving registration details: ${err}`);
      res.status(500).send({ message: "Some error occurred while retrieving registration." });
    } else {
      logger.info(`Successfully retrieved registration details: ${JSON.stringify(data.Items)}`);
      res.send(data.Items); // DynamoDB returns results in the Items array
    }
  });
};

Key Notes to Remember #

• Skip scan entirely: A scan reads every item in the table, which is slow and expensive for large datasets. Using a GSI with query is the optimal approach here.
• Match data types: Ensure the data type of requester_manager_id in the index matches what's stored in the table (e.g., string vs. number). Mismatched types will cause the query to return no results.
• Optimize index projection: If you don't need all attributes, use ProjectionType=INCLUDE to only include the fields you need in the index. This reduces storage costs and improves query performance.

内容的提问来源于stack exchange，提问作者Shivakumar Gadade