Linux send queue is not zero, how to diagnose?

Question

I used redis-cli to connect to redis-server, which just stuck forever.

I used strace to trace the redis-cli process and found this:

[root@iZ2ze0fgielr9er134vdz0Z zuoca]# strace -p 8044
strace: Process 8044 attached
read(3,

Also, I used netstat to check the corresponding connection status and found this:

[root@iZ2ze0fgielr9er134vdz0Z zuoca]# netstat -pan | grep 6379
tcp        0      0 10.85.125.2:6379        0.0.0.0:*               LISTEN      29628/redis-server
tcp        0      0 10.85.125.2:63576       10.85.125.2:6379        ESTABLISHED 8044/redis-cli
tcp        0    327 10.85.125.2:6379        10.85.125.2:63576       ESTABLISHED 29628/redis-server
[root@iZ2ze0fgielr9er134vdz0Z zuoca]#

You can see that the send queue is not zero while redis-cli process stuck there to wait to read something, it's so weird.

I've tried to start a new Redis server instance, still can not connect to it.

Redis version info:

Redis server v=3.2.12 sha=00000000:0 malloc=jemalloc-3.6.0 bits=64 build=7897e7d0e13773f