Open nqb opened 2 years ago
Additional notes:
http://IP_OF_PORTAL/captive-portal
on management interface with portal
daemon enabled. I doesn't work when trying to reach captive portal on a registration interface (case described above).haproxy-portal
requests:
# on a registration interface
Aug 23 09:52:21 pfdeb11dev haproxy-portal-docker-wrapper[1131]: 172.18.201.115:46338 [23/Aug/2022:09:52:21.053] portal-http-172.18.200.12 proxy/proxy 0/0/0/5/5 200 602 - - ---- 2/1/0/0/0 0/0 {172.18.201.2} "GET /captive-portal HTTP/1.1"
# on management interface
Aug 23 09:52:27 pfdeb11dev haproxy-portal-docker-wrapper[1131]: 172.18.200.251:41690 [23/Aug/2022:09:52:27.534] portal-http-172.18.200.12 172.18.200.12-backend/containers-gateway.internal:8080 0/0/0/50/50 200 4916 - - ---- 2/1/0/0/0 0/0
{172.18.200.12} "GET /captive-portal HTTP/1.1"
I hit same issue while running inline
tests.
Still unable to get the captive portal working with an Aruba2930M Series switch. Clients in the registration VLAN are redirected to msftconnecttest.com/redirect and after a while it redirects to the pfserver https://packetfence.nac/captive-portal?destination_url=http://www.msftconnecttest.com/redirect. my clients in registration can ping packetfence.nac or any other URL because it's redirected to the IP 66.70.255.147. They can also ping all the packetfence interfaces IP's.
Describe the bug I noticed a difference when reaching captive portal through IP address (not using name) in Venom tests.
Previously, in order to check locales on captive portal, I used following URL:
https://172.18.201.2/captive-portal
which returned directly captive portal webpage. Now, it seems that this URL returned:If I reached directly
https://pf.example.lan/captive-portal
(withpf.example.lan
configured as general domain and hostname inpf.conf
) , I got captive portal webpage.I just opened that issue to discuss about difference and potential impact it will have.