During the process of changing a lot of network configuration last weekend, I upgraded from RouterOS 5.6 to 5.7 on an RB1100. On Monday I discovered that DHCP had stopped working.
DHCP requests were being seen by the RB1100 interfaces (using Torch), and the RB1100 could see DHCP traffic on the network (it detected a rogue DHCP server we set up for testing using an alert), it wasn’t firewall rules, there was nothing that could sensibly explain why the RB1100 simply refused to respond to any of the requests with an offer.
After a further day of trying to figure out which of the the many network configuration changes made during the weekend could have possibly caused it, I’ve just downgraded back to 5.6 from 5.7 and DHCP has suddenly sprung back to life. Configuration has remained the same.
Has something changed in 5.7 that could possibly cause DHCP to break using an identical configuration? The only talk I’ve seen about DHCP changes have been to get ready to introduce DHCPv6 but that’s not until 5.8, and it can’t be completely broken - our rogue DHCP server was running on an SXT, also on 5.7 (albeit mips rather than ppc).