Last checked and updated on 13 Jun 2026
What Is New in Redis 1.3 (summary table)
Redis 1.3 introduces foundational improvements focused on data persistence and client handling, setting the stage for its role as a high-performance store.
| Category | Key Changes |
|---|---|
| New Features | Background saving, Append-only file (AOF) persistence |
| Improvements | Better handling of many connected clients, Optimized memory usage for lists |
| Bug Fixes | Fixes for replication and protocol parsing |
| Internal | Refactored networking code, More efficient data structure operations |
How did persistence evolve in Redis 1.3?
The biggest shift is the move towards more robust and flexible data durability. Before, saving a snapshot (RDB) would block all operations. Now, Redis 1.3 can perform saves in the background using a fork, keeping the server responsive.
This version also introduces the append-only file (AOF) persistence mode. Instead of periodic snapshots, you can log every write command to a file. This matters because it offers finer granularity for recovery, minimizing data loss between saves.
The Impact on Operations
In practice, background saving means no more client timeouts during a SAVE on large datasets. The AOF file is a game-changer for use cases where data integrity is critical, though it comes with a trade-off in disk I/O and file size.
What changes were made for client and memory efficiency?
Redis 1.3 significantly improved how it manages many concurrent connections. The networking layer was refactored to be more scalable, reducing the overhead per client. This is core to Redis's identity as a fast, concurrent data server.
Memory usage for list data structures was optimized. Internally, the way lists are stored and manipulated became more efficient, which directly translates to lower memory footprint and faster operations for applications using lists heavily.
Were there important fixes for replication?
Yes, several bugs related to the master-slave replication flow were addressed. These fixes improved the stability and reliability of data synchronization between nodes, which is essential for building basic high-availability setups.
Protocol parsing edge cases were also corrected. This ensures better compatibility with various client libraries and prevents unexpected disconnections or errors during command processing.
FAQ
Does background saving completely replace the old SAVE command?
No, the synchronous SAVE command still exists. The new BGSAVE command triggers the background save. Use SAVE only in specific scenarios where you need a guaranteed immediate snapshot and can afford the block.
Should I use RDB snapshots or the new AOF persistence?
It depends on your data loss tolerance. RDB is faster and creates compact snapshots but can lose data since the last save. AOF logs every write, offering better durability at the cost of slower writes and larger files. You can use both.
How does the AOF file work with log rotation?
Redis can rewrite the AOF file in the background to compact it, removing redundant operations. You can also manually send a BGREWRITEAOF command. This is similar to how BGSAVE works but for the append-only log.
Do I need to change my client code for Redis 1.3?
Most likely not. The wire protocol remains compatible. The changes are mainly server-side for persistence, memory, and connection handling. Existing clients should work without modification.
What's the real-world performance impact of the networking refactor?
You'll see the benefit when handling thousands of idle or active connections. The server uses fewer system resources per connection, allowing a single instance to handle more concurrent clients smoothly, which is crucial for scaling.