Prompt
Redis keeps going down on me
Latest observation
Oct 25, 2026
OpenAI APIWeb search: off
Sorry you’re dealing with that — Redis “going down” can mean a few different things, and the fix depends on why it’s stopping.
First: what exactly is happening?
Common cases:
- Redis process exits/crashes
- Redis becomes unreachable but the process is still running
- Redis gets killed by the OS (often OOM / memory pressure)
- Redis is evicted/restarted in Docker/Kubernetes
- It’s overloaded and timing out
Quick checks
Run these when it happens:
1) Check Redis logs
sudo journalctl -u redis --since "1 hour ago"
or wherever your Redis log file is:
tail -n 200 /var/log/redis/redis-server.log
2) Check if the OS killed it
dmesg -T | grep -i -E 'killed process|oom'
If you see OOM kills, Redis is using too much memory or the machine is under pressure.
3) Check Redis status
redis-cli ping
redis-cli info memory
redis-cli info server
4) Check system resources
free -h
top
df -h
Very common causes
Memory issues
- Redis has no memory cap
- RDB/AOF background saves fail
- The server runs out of RAM and the kernel kills Redis
Useful config:
maxmemory 2gb
maxmemory-policy allkeys-lru
Persistence problems
If Redis is configured to save snapshots or use AOF:
- disk full
- permission issues
- corrupted AOF/RDB
- background save failures
Check:
redis-cli info persistence
Docker/Kubernetes restarts
- container memory limit too low
- liveness probe too aggressive
- pod rescheduled
Network / connection issues
If apps say Redis is “down” but server is fine:
- firewall
- binding only to localhost
- wrong
protected-mode - server listening on a different interface/port
Check:
ss -ltnp | grep 6379
If you want the fastest path to a diagnosis
Send me:
- The last 50 lines of Redis logs
- Output of:
redis-cli info memory redis-cli info persistence - Whether this is on bare metal, VM, Docker, or Kubernetes
- What “down” means: process exits, connection refused, timeout, or crash
If you paste the log snippet here, I can help pinpoint the cause quickly.