如何在Django中将一个项目的完整数据复制到另一个项目？

Hey Ryan, great question—you absolutely don’t have to resort to manual copy-pasting here. Since your new project uses an identical data model, there are several reliable, efficient ways to migrate your entire Postgres dataset (including those private user posts) from your old EC2 instance to the new one. Let’s walk through the most practical options:

1. Postgres Native pg_dump + pg_restore (Most Common Approach) #

This is the go-to method for most Postgres migrations, especially when your database isn’t extremely large.

Step 1: Create a Full Backup of the Old Database #

On your old EC2 instance, log into the Postgres user (usually postgres) and run the pg_dump command to export the entire database:

pg_dump -U your_postgres_username -d old_database_name -F c -b -v -f /tmp/old_db_backup.dump

Let’s break down the flags:

• -F c: Uses a custom compressed format (faster to transfer and restore)
• -b: Includes large objects (critical if your private content has attachments like images or documents)
• -v: Runs in verbose mode so you can track the backup progress
• -f: Specifies the output file path

Step 2: Transfer the Backup to the New EC2 Instance #

Use scp (secure copy) to move the backup file from the old instance to the new one. Replace the placeholders with your actual EC2 details:

scp /tmp/old_db_backup.dump ec2-user@your-new-ec2-ip:/tmp/

If you prefer using AWS S3 for safer transfer (great if instances are in different regions or you want a backup stored in S3), you can upload the dump to S3 first:

# On old EC2:
aws s3 cp /tmp/old_db_backup.dump s3://your-backup-bucket-name/

# On new EC2:
aws s3 cp s3://your-backup-bucket-name/old_db_backup.dump /tmp/

Step 3: Restore the Backup to the New Database #

First, make sure your new database is already set up (since your model is identical, you should have the same table structure in place). If you haven’t created the new database yet:

createdb new_database_name -U your_postgres_username

Then run pg_restore to import the data:

# If you want to overwrite existing table structures (only do this if the new DB is empty):
pg_restore -U your_postgres_username -d new_database_name -v /tmp/old_db_backup.dump

# If the new DB already has the correct table structure (recommended), use `-a` to only import data:
pg_restore -U your_postgres_username -d new_database_name -a -v /tmp/old_db_backup.dump

2. Postgres Streaming Replication (For Large Databases or Minimal Downtime) #

If your database is large (100GB+) or you need to minimize downtime during migration, streaming replication is a better option. It lets you sync the new instance with the old one in real-time, then switch over once sync is complete.

High-Level Steps: #

1. On the old EC2 instance’s postgresql.conf, update these settings:

   wal_level = replica
   max_wal_senders = 5  # At least 1, 5 gives room for future replication
   wal_log_hints = on
   

2. Edit pg_hba.conf on the old instance to allow the new EC2 instance’s IP to connect for replication:

   host    replication     replication_user    your-new-ec2-ip/32    scram-sha-256
   

   (Create a dedicated replication user if you don’t have one already.)
3. Restart Postgres on the old instance to apply changes.
4. On the new EC2 instance, stop Postgres, clear the default data directory, and run a base backup from the old instance:

   sudo systemctl stop postgresql
   sudo rm -rf /var/lib/postgresql/14/main/*  # Replace 14 with your Postgres version
   pg_basebackup -h old-ec2-ip -U replication_user -D /var/lib/postgresql/14/main -P -X stream
   

5. Configure the new instance to act as a replica, then start Postgres. Once replication catches up, you can promote it to the primary database and point your new app to it.

Critical Pre-Migration Checks #

• Pause Writes to the Old App: Before starting any migration, put your old beta app into read-only mode (or shut it down temporarily) to prevent new data from being added during the transfer—this ensures your new database has a consistent snapshot.
• Verify Postgres Version Compatibility: Make sure the Postgres version on the new EC2 instance is the same as the old one (or at least compatible; newer versions can usually restore backups from older ones, but not vice versa).
• Validate Migrated Data: After restoration, spot-check a few private posts, compare row counts for key tables, and run any app-specific tests to ensure all data made it over correctly.

内容的提问来源于stack exchange，提问作者Ryan Oh