Fix recurring errors when using TLS/SSL

Turns out haproxy needs to be told that the connection is TLS/SSL twice.
Once for the `tcp-check` connect and then for the `server-template`.

If `ssl` is not specified in the `server-template` config, then there
will be some plaintext tcp connections. These seem to be separate from
the `tcp-check` connections though.

Redis will print an error such as this when it's running in TLS/SSL mode
and a plaintext connection gets opened:

```
1:M 14 Apr 2025 14:41:03.686 # Error accepting a client connection: error:0A00010B:SSL routines::wrong version number (addr=11.35.244.11:34886 laddr=11.35.244.4:6379)
```

They are mostly annoying, but are harmless.

Author: Kidswiss

appuio/haproxy-2.5.1.tgz

@@ -56,5 +56,5 @@ data:
       tcp-check expect string +OK
       {{- end }}
 
-      server-template redis- {{ $redisk8s.nodeCount }} {{ $redisk8s.dnsservicename }}:{{ default "6379" $redisk8s.port }} {{ if $redisk8s.check.enabled }}check inter 1s on-marked-down shutdown-sessions {{ end }}resolvers mydns init-addr none {{ if $redisk8s.tls }} verify none {{ end }}
+      server-template redis- {{ $redisk8s.nodeCount }} {{ $redisk8s.dnsservicename }}:{{ default "6379" $redisk8s.port }} {{ if $redisk8s.check.enabled }}check inter 1s on-marked-down shutdown-sessions {{ end }}resolvers mydns init-addr none {{ if $redisk8s.tls }} verify none ssl {{ end }}
 {{- end }}