03 / 04

What is AOF rewrite, and what problems can it cause in production?

Difficulty: 6/10
AOF rewrite, performance impact, disk I/O

AOF rewrite compacts the log by replaying the current state. It forks a child process, which triggers copy-on-write (CoW). On a write-heavy instance with large memory, this can spike memory usage (up to 2x) and increase latency as the OS copies dirty pages.

Scenario Questions

0-2 years experience

  1. 1If you enable AOF persistence and the background rewrite process starts, what will you see in the Redis log, and how does it affect write latency for a client issuing SET commands?
  2. 2You have a small Redis instance with a 500 MB AOF file. How would you manually trigger an AOF rewrite, and what steps would you take to verify it completed successfully?
  3. 3What happens if the AOF rewrite process runs out of disk space on the server?

2-5 years experience

  1. 1During a load test, you notice latency spikes when the AOF rewrite kicks in. Walk me through how you would diagnose the root cause and what configuration knobs you might adjust.
  2. 2A recent deployment caused the Redis AOF rewrite to fail repeatedly, leading to a growing AOF file. Explain how you would troubleshoot the failure and prevent data loss.
  3. 3Explain the trade‑offs between using appendonly yes with auto‑aof‑rewrite‑percentage versus scheduling manual rewrites via BGREWRITEAOF in a production service.

5-8 years experience

  1. 1Design a strategy to minimize the impact of AOF rewrites on a high‑throughput write‑heavy Redis cluster that must stay under a 5 ms latency SLA.
  2. 2How would you handle a scenario where an AOF rewrite consumes all available memory on a Redis instance, causing the process to be killed? Discuss mitigation and architectural changes.
  3. 3Compare the implications of AOF rewrite versus RDB snapshots for disaster recovery in a multi‑region deployment, focusing on recovery time objectives and operational risk.

8+ years experience

  1. 1Your organization plans to migrate a legacy monolithic service using Redis AOF persistence to a microservices architecture with separate read/write replicas. How would you redesign the persistence strategy to avoid AOF rewrite bottlenecks while ensuring durability?
  2. 2At scale, frequent AOF rewrites can cause I/O contention across many nodes. Propose a cross‑team policy and tooling approach to coordinate rewrite windows, monitoring, and fallback mechanisms.
  3. 3Discuss the long‑term maintenance considerations of keeping AOF rewrite enabled in a system that must retain ten years of audit logs. What architectural patterns could you employ to balance storage cost, compliance, and performance?

Follow-up Questions

  • What metrics would you monitor while an AOF rewrite is in progress?
  • How do you decide appropriate values for auto‑aof‑rewrite‑percentage and auto‑aof‑rewrite‑min‑size?
  • If a rewrite fails, what steps do you take to ensure data durability?
Share

Share via WhatsApp, X, Facebook, LinkedIn or copy link. Open Graph preview enabled.