there seems to be an issue with IPv6 on the japan node. According to mtr there is 100% packet loss after the first hop. Is this already known?
It did work until i reconfigured the network because i saw there was a different and bigger subnet in the solus panel.
[email protected]:~# mtr -6 --report-wide 2001:4860:4860::8888 Start: Mon Mar 20 20:23:11 2017 HOST: athos Loss% Snt Last Avg Best Wrst StDev 1.|-- 2001:19f0:7000:8005:5054:ff:fe1c:28ae 0.0% 10 0.4 0.4 0.4 0.7 0.0 2.|-- ??? 100.0 10 0.0 0.0 0.0 0.0 0.0
Yes, it is the right IPv6 address. I even reconfigured the network through solus panel. I also reinstalled the whole vm just to try it out. Same result.
You can check for your self (same signup email as here on the board) if you want, nothing private is on the server at the moment. So either there is a network problem or i'm blind and cant see my error.
I will have a look, but that fact that it was working until you made vps specific changes tells me it is vps specific.
Ah, interesting, I see the same symptom on a VPS that hasn't been reconfigured for ages. Centos 7 (don't ask).
It looks as though I've been allocated a new separate IPv6 /80 subnet at some time which I've never configured, but my original IPv6 address from a different subnet is clearly still mapped for inbound traffic, I can/do ssh in on IPv6, but can't connect out (as I'd expect).
Not an issue for me but thought I'd mention it, I don't recall seeing anything about a change.
EDIT: Spoke to soon, I do seem to have the /80 subnet configured, so yes I experience this issue too, though I've always connected on an IPv6 address that no longer appears in the admin.
Last edited by cochon (2017-03-21 11:44:18)
You had an old IPv6 subnet configured that was issues about a year ago, I have switched it back to the correct range and IPv6 works fine now.
I will probably be moving everything to Linode anyway in the near future so everyone will get a bigger IPv6 range.