I have an OpenBSD box hosted with Bytemark. I recently updated it to the latest, OpenBSD 6.3. After the update, my IPv6 networking stopped working.
This is a documented change in the upgrade notes, but I guess I didn't read close enough,
or didn't realise the impact this change would have. Hopefully noting it here might be useful for anyone else that spends half an hour scratching
their head wondering what happened.
After doing the upgrade all my IPv6 stopped working, and I could no longer ping the default gateway on it's linklocal address, fe80::1%vio0.
The workaround is simple, just add -soii to your hostname.if file in order to restore the previous behaviour of addresses generated from the MAC address.
After restarting the networking with sh /etc/netstart vio0 I'm able to ping again:
traci# ping6 fe80::1%vio0
PING fe80::1%vio0 (fe80::1%vio0): 56 data bytes
64 bytes from fe80::1%vio0: icmp_seq=0hlim=64time=12.483 ms
64 bytes from fe80::1%vio0: icmp_seq=1hlim=64time=0.674 ms
64 bytes from fe80::1%vio0: icmp_seq=2hlim=64time=0.772 ms
64 bytes from fe80::1%vio0: icmp_seq=3hlim=64time=1.074 ms
--- fe80::1%vio0 ping statistics ---
4 packets transmitted, 4 packets received, 0.0% packet loss
round-trip min/avg/max/std-dev = 0.674/3.751/12.483/5.044 ms
I'm guessing that as this is a virtual server, perhaps the underlying hypervisor/router is expecting a particular address.
This is a talk given at Codemobile 2017 conference in Chester, UK. It was a 5 minute lightning
talk detailing how to automate the building of apps for each and every feature branch created
as part of a git-flow workflow.