[ClusterLabs Developers] Unable to access Upstream Redis port using Virtual IP

[Ken Gaillot]

Hi,

This would be better for the users at clusterlabs.org list, where people
have experience with all sorts of setups. This list is just for
discussing code.

Thanks!

On Mon, 2022-04-25 at 13:23 +0530, Michel D wrote:
> I am running two nodes(172.16.10.52/172.16.10.53) NGINX load
> balancers. The load balancers are configured to proxy pass upstream
> Redis server using TCP port.
> 
> stream {
> 
>         upstream redis_cache {
>                                 server 172.16.10.242:6379 max_fails=3
> fail_timeout=30s;
>                                 server 172.16.10.48:6379 max_fails=3
> fail_timeout=30s;
>                                 server 172.16.10.49:6379 max_fails=3
> fail_timeout=30s;
>         }
> 
>         server {
>                   listen 172.16.10.53:6379;
>                   proxy_pass redis_cache;
>         }        
> }
> 
> The above setting works fine for both the load balancers. I can
> access the upstream redis server using load balancers/redis-cli.
> 
> # redis-cli -h 172.16.10.53 -p 6379 ---> Works fine
> 
> Now I have set up a  corosync/pacemaker cluster including load
> balancer 1(172.16.10.52) and load balancer 2(172.16.10.53) to access
> the upstream redis server using virtual IP(172.16.10.51)
> 
> I can access the cluster using the virtual IP(http://172.16.10.51)
> serving NGINX default welcome page. 
> 
> However, I am unable to access the upstream redis server using Redis-
> cli/Virtual IP.
> I am getting a Connection refused error.
> 
> # redis-cli -h 172.16.10.51 -p 6379
> Could not connect to Redis at 172.16.10.51:6379: Connection refused
> Could not connect to Redis at 172.16.10.51:6379: Connection refused
> 
> Steps used to create the cluster
> ************************************
> 
> # pcs cluster auth redis-lb1 redis-lb2 -u hacluster -p welcome --
> force
> 
> # pcs cluster setup --force --name pacemaker1 redis-lb1 redis-lb2
> 
> # pcs cluster start --all
> 
> # pcs property set stonith-enabled=false
> # pcs property set no-quorum-policy=ignore
> # pcs status
> 
> # pcs resource create virtual_ip ocf:heartbeat:IPaddr2
> ip=172.16.10.51 cidr_netmask=32 op monitor interval=10s
> 
> # pcs resource create nginx-lb ocf:heartbeat:nginx
> configfile=/etc/nginx/nginx.conf op monitor timeout="5s"
> interval="5s"
> 
> # pcs constraint colocation add nginx-lb virtual_ip INFINITY
> 
> # pcs constraint order virtual_ip then nginx-lb
> 
> # pcs constraint location nginx-lb prefers redis-lb1=50
> # pcs constraint location nginx-lb prefers redis-lb2=50
> 
> # pcs cluster stop --all
> 
> # pcs cluster start --all
> 
> ********************************
> Shall i have to provide the port number of  Redis configuration(6379)
> that was setup in load balancers while creating pcs resource creation
> steps ?
> 
> Regards
> _______________________________________________
> Manage your subscription:
> https://lists.clusterlabs.org/mailman/listinfo/developers
> 
> ClusterLabs home: https://www.clusterlabs.org/
-- 
Ken Gaillot <kgaillot at redhat.com>