auto_eject_host not working.

[leunamnauj]

I have 3 redis servers + sentinel and I thought on put twemproxy over for load balancing and HA somehow.
I understand that auto_eject_host directive is used to remove a server from the pool when is failing but this is not the case, when I put down one of Redis servers twemproxy still trying to query on it and I got "(error) ERR Connection refused"

Honestly I'm not sure if I set something wrong of if I misunderstand the behavior of the tool.
Thanks in advance

This is my config:

redis1:
listen: 0.0.0.0:6379
redis: true
hash: fnv1a_64
distribution: ketama
auto_eject_hosts: true
timeout: 400
server_retry_timeout: 2000
server_failure_limit: 1
preconnect: true
servers:

• privip1:6380:1
• privip1:6380:1
• privip1:6380:1

[charsyam]

hi, @leunamnauj
auto_eject_hosts works like below:

1. eject server.
2. after sometime add it again without checking
3. loop 1

so, I think you met this kind of situation.
If you want checking of conneciton.
try this: #428

in mycase, It works fine.

[yumitubi]

@charsyam thank you for your patch!

Do you use this changes in production? Can I add this code to my installation of twemproxy?