Absolutely—there are several rock-solid methods and tools for backing up and restoring Redis, perfect for your scenario of moving from an outdated Windows Redis instance to a production Linux setup. Let’s break down the most trustworthy options tailored to your needs:

1. RDB Snapshot (Official, Full Backup) #

This is the most straightforward and widely recommended approach, especially if you can tolerate a short service downtime during migration.

Steps: #

• Trigger a manual snapshot on your Windows Redis:
  • Use BGSAVE (runs in the background, no service interruption) if your instance is still serving traffic.
  • Use SAVE (blocks the server until the snapshot is done) only if the instance is already offline.
• Locate the RDB file: Run CONFIG GET dir and CONFIG GET dbfilename in your Windows Redis CLI to find the exact path (usually something like C:\Program Files\Redis\dump.rdb for the official Microsoft Archive build).
• Secure the file: Stop the Windows Redis service to prevent any writes that could corrupt the snapshot, then copy dump.rdb to your Linux server (use SCP, SFTP, or a file transfer tool of your choice).
• Restore on Linux:
  1. Run CONFIG GET dir on your Linux Redis to get its data directory (typically /var/lib/redis/).
  2. Move the copied dump.rdb into this directory, ensuring the filename matches the dbfilename setting in your Linux Redis config.
  3. Start the Linux Redis service—it will automatically load the RDB file on startup.
• Verify: Use commands like DBSIZE to check key counts, or sample critical keys with GET/HGETALL to confirm data integrity.

2. AOF Persistence (Real-Time, Incremental Backup) #

If your Windows Redis has Append-Only File (AOF) enabled, this method gives you a more up-to-date backup (since it logs every write operation).

Steps: #

• Optimize the AOF file: Run BGREWRITEAOF on your Windows Redis to compact the file (removes redundant operations and reduces size).
• Secure and transfer: Stop the Windows service, then copy appendonly.aof to your Linux Redis data directory.
• Restore on Linux:
  1. Edit your Linux Redis config to set appendonly yes and ensure appendfilename matches the copied file name.
  2. If the AOF file is corrupted, run redis-check-aof --fix /path/to/appendonly.aof to repair it.
  3. Start the Linux Redis service—it will replay the AOF log to restore all data.

3. Online Migration Tools (Minimize Downtime) #

For large databases where downtime isn’t acceptable, use dedicated migration tools that support full + incremental sync:

Redis-Shake #

A popular open-source tool that handles both full RDB snapshots and real-time incremental sync via the Redis replication protocol.

• Setup: Deploy Redis-Shake on your Linux server (or a middle machine).
• Configure: Point it to your Windows Redis (source) and Linux Redis (target) endpoints.
• Run: Start with a full sync to copy existing data, then switch to incremental sync to capture new writes. Once sync is complete, cut over traffic to the Linux instance.

redis-cli MIGRATE (For Targeted Key Migration) #

While you mentioned this, it’s better suited for migrating specific keys rather than an entire database. For full migration, it’s less efficient than RDB or Redis-Shake, but here’s how to use it in bulk:

redis-cli --source windows-redis:6379 --target linux-redis:6379 --keys '*' --copy --replace

Note: This can be slow for large datasets, and doesn’t handle ongoing writes during migration.

Critical Notes for Your Migration #

• Version Compatibility: Old Windows Redis versions may use older RDB/AOF formats. Test loading the backup in a staging Linux Redis instance first to avoid compatibility errors.
• Data Validation: Always compare key counts (DBSIZE) and sample critical data before and after migration to ensure nothing is lost.
• Permissions: On Linux, make sure the redis user has read/write access to the data directory and the copied backup files.
• Downtime Planning: If using RDB/AOF, schedule the migration during a low-traffic window to minimize impact.

内容的提问来源于stack exchange，提问作者patachi