There is a hotspot with 192.168.130.0/23 on hotspot interface (internal WiFi bridged together with 4 LAN ports with another 6 WiFi APs) which uses DHCP relay from the main router. It works, but some clients complain that they can’t connect to WiFi. I found a problem, here is an example from the main router’s log:
- 08:12:46 dhcp,info dhcp-hotspot-130 deassigned 192.168.130.120 from 1C:XX:XX:XX:XX:1C
- 08:18:06 dhcp,info dhcp-hotspot-130 assigned 192.168.130.228 to 1C:XX:XX:XX:XX:1C
- 08:35:43 dhcp,info dhcp-hotspot-130 deassigned 192.168.130.228 from 1C:XX:XX:XX:XX:1C
- 08:40:15 dhcp,info dhcp-hotspot-130 assigned 192.168.130.120 to 1C:XX:XX:XX:XX:1C
Well, step 1 says that IP .130.120 was deassigned. But I don’t understand why few minutes later another IP .130.228 was assigned (step 2) and then again - it was deassigned (step 3) and few minutes later the first IP (.130.120) was assigned. Shouldn’t DHCP offer the same IP? DHCP lease time is set to 7 days. Strange thing is that this only appears for DHCP relays, not for normal DHCP clients. Could somebody explain it?
It may look like there is no problem, but it makes problem with hotspot autentification - as there are two records in hotspot hosts table, it means that two sessions of UserManager are used instead of one and it leads to “simultaneous session limit reached”. I don’t want to increase session limit, I’d like to prevent creating needless sessions.
There is ROS 5.22 on hotspot (RB751G) and ROS 5.21 on the main router (RB1200).