'Stale socket in lsof due to REDIS
I'm having a redis connection established and I've a python script which performs CLUSTER MEET, CLUSTER FORGET, IPTBL ADD and IPTBL DELETE.
When I give CLUSTER FORGET. Before FIN packet reaches the other end, it closes the connection and I'm left with a stale socket.
On going through the redis documentation and other blogs I found tcp-keepalive and even after changing it to 300 and restarting the redisd I still see no effect.
/usr/sbin/lsof | grep 16381
redis-ser 23104 admin 10u IPv6 397237 0t0 TCP *:16381 (LISTEN)
redis-ser 23104 admin 11u IPv4 397239 0t0 TCP *:16381 (LISTEN)
redis-ser 23104 admin 12u IPv4 397251 0t0 TCP a.b.c.d:16381->a.b.c.e:56440 (ESTABLISHED)
redis-ser 23104 admin 13u IPv4 397253 0t0 TCP a.b.c.d:16381->a.b.c.f:58744 (ESTABLISHED)
redis-ser 23104 admin 14u IPv4 408062 0t0 TCP a.b.c.d:47230->a.b.c.e:16381 (ESTABLISHED)
redis-ser 23104 admin 15u IPv4 748468 0t0 TCP a.b.c.d:39754->a.b.c.f:16381 (ESTABLISHED)
redis-ser 23104 admin 16u IPv4 719475 0t0 TCP a.b.c.d:16381->a.b.c.f:58868 (ESTABLISHED)
I would like to understand how to make tcp-keepalive work ? Am I missing something ?
Any help would be appreciated.
Sources
This article follows the attribution requirements of Stack Overflow and is licensed under CC BY-SA 3.0.
Source: Stack Overflow
| Solution | Source |
|---|