生产环境下如何将ServiceSchema的vehicleType迁移至CustomerSchema并同步数据

Got it, let's walk through this step by step since you're dealing with a production environment serving 400+ customers—we need to make sure this migration is safe, complete, and doesn't disrupt your service.

First, modify your CustomerSchema to include the vehicleType field (match the same String type as it was in serviceSchema):

var CustomerSchema = new Schema({
  name:{ type: String },
  phone:{ type: Number },
  vehicleNo:{ type: String },
  vehicleType:{ type: String } // Add this new field to store vehicle type
})

Deploy this schema update to your production environment first. Mongoose will handle the new field gracefully, allowing you to start writing data to it without breaking existing functionality.

You'll need to copy the vehicleType value from each Service document to its associated Customer. Below are two options based on your preferred tooling:

Option A: Node.js/Mongoose Script (Aligned with Your Stack) #

This script iterates through all services, maps them to customers, and syncs the field. It also handles edge cases like missing data or conflicting values:

const mongoose = require('mongoose');
const Customer = require('./models/Customer'); // Path to your Customer model
const Service = require('./models/Service'); // Path to your Service model

async function migrateVehicleType() {
  try {
    // Connect to your MongoDB instance
    await mongoose.connect('your-mongodb-connection-string', {
      useNewUrlParser: true,
      useUnifiedTopology: true
    });

    // Use a cursor to process services in batches (gentler on memory)
    const serviceCursor = Service.find().cursor();
    
    for await (const service of serviceCursor) {
      // Skip invalid records
      if (!service.customer || !service.vehicleType) {
        console.log(`Skipping service ${service._id}: missing customer or vehicleType`);
        continue;
      }

      // Find the linked customer
      const customer = await Customer.findById(service.customer);
      if (!customer) {
        console.log(`Customer ${service.customer} not found for service ${service._id}`);
        continue;
      }

      // Sync the vehicleType only if the customer doesn't have one yet
      if (!customer.vehicleType) {
        customer.vehicleType = service.vehicleType;
        await customer.save();
        console.log(`Updated customer ${customer._id} with vehicleType: ${service.vehicleType}`);
      } else if (customer.vehicleType !== service.vehicleType) {
        // Handle conflicts: same customer has services with different vehicle types
        console.warn(`Conflict: Customer ${customer._id} has vehicleType ${customer.vehicleType}, but service ${service._id} has ${service.vehicleType}. Review manually.`);
      }
    }

    console.log('Migration completed successfully!');
    process.exit(0);
  } catch (err) {
    console.error('Migration failed:', err);
    process.exit(1);
  }
}

migrateVehicleType();

Option B: MongoDB Shell Script #

If you prefer using the MongoDB shell directly, this script achieves the same result:

// Switch to your database
use your-production-db;

// Iterate through all service documents
db.services.find().forEach(function(service) {
  if (!service.customer || !service.vehicleType) {
    print(`Skipping service ${service._id}: missing customer or vehicleType`);
    return;
  }

  const customer = db.customers.findOne({ _id: service.customer });
  if (!customer) {
    print(`Customer ${service.customer} not found for service ${service._id}`);
    return;
  }

  if (!customer.vehicleType) {
    db.customers.updateOne(
      { _id: customer._id },
      { $set: { vehicleType: service.vehicleType } }
    );
    print(`Updated customer ${customer._id} with vehicleType: ${service.vehicleType}`);
  } else if (customer.vehicleType !== service.vehicleType) {
    print(`Conflict: Customer ${customer._id} has vehicleType ${customer.vehicleType}, but service ${service._id} has ${service.vehicleType}. Review manually.`);
  }
});

Don't skip this step! Confirm that all data was synced correctly:

• Run a query to check for customers missing vehicleType (should only include customers with no associated services):

  db.customers.find({ vehicleType: { $exists: false } });
  

• Check for mismatched values between services and their linked customers:

  db.services.aggregate([
    {
      $lookup: {
        from: "customers",
        localField: "customer",
        foreignField: "_id",
        as: "customer"
      }
    },
    { $unwind: "$customer" },
    {
      $match: {
        $expr: { $ne: ["$vehicleType", "$customer.vehicleType"] }
      }
    }
  ]);
  

Now that data is migrated, adjust your codebase:

• Read from Customer: All places where you previously fetched service.vehicleType should now use customer.vehicleType
• Write to Customer: Ensure new customer creation/updates set vehicleType correctly
• Deprecate Service Field: Keep serviceSchema.vehicleType temporarily (in case of old code versions running), but mark it as deprecated. After confirming all code is updated, you can remove this field from the schema and database.

• Backup First: Take a full backup of your customers and services collections before running the migration
• Test in Staging: Run the migration script in a staging environment first to catch issues
• Low Traffic Window: Execute the migration during a period of low user activity to minimize impact
• Monitor Logs: Keep an eye on migration logs to address errors or conflicts immediately
• Rollback Plan: If something goes wrong, restore from your backup and revert the schema changes

内容的提问来源于stack exchange，提问作者foxenn_pro