503 if backend node goes down

I have following configuration,

global
log 127.9.9.1 local0 #debug
log /dev/log local0
log /dev/log local1 notice
chroot /var/lib/haproxy
stats socket /run/adminha.sock mode 600 level admin
stats timeout 30s
user haproxy
group haproxy
daemon

defaults
log global
mode http
option httplog
timeout connect 5000
timeout client 2m
timeout server 2m

frontend outwards-haproxy
bind *:80
bind *:443 ssl crt /pathgiven no-sslv3
redirect scheme https code 301 if !{ ssl_fc }
stats enable
stats uri /stats
stats enable
stats uri /stats
stats hide-version
stats auth admin:1234

acl is_off url_beg /smoke1/
use_backend offapp if is_off

backend offapp
balance roundrobin
option redispatch 1
retries 2
cookie platform_extension_sessionid prefix nocache
server off1 host1:8443 check ssl verify none cookie s1
server off2 host2:8443 check ssl verify none cookie s2

I m getting 503 if one of backend server is stopped. Why it is not seamlessly transferring connection ?? even if i have 2 nodes