Users are starting to ask if it's possible to reach libravatar.org via IPv6 as well and I'd love to solve that. libravatar.org is a VM somewhere in Fedoras Infra and I'd be glad if someone could provide an IP.
I'm currently only partially available, but I should be back on 25th of March and with full capacity, so if you could provide the IP until then and let me know what I need to configure, that would be awesome.
Metadata Update from @zlopez: - Issue priority set to: Waiting on Assignee (was: Needs Review) - Issue tagged with: medium-gain, medium-trouble, ops
Metadata Update from @kevin: - Issue assigned to kevin
ok, this is a aws instance.
I set the subnet it's on to auto assign ipv6 addresses. It however doesn't have one right now. You may need to down/up the interface or reboot the instance to get a ipv6 address assigned?
Thanks for that Kevin! Seems there is something else we need to do - rebooted already and set ipv6 method to auto.
$ ip a show eth0 2: eth0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 9001 qdisc mq state UP group default qlen 1000 link/ether 02:fa:d8:6a:0a:e1 brd ff:ff:ff:ff:ff:ff altname enp0s5 altname ens5 inet 172.30.5.38/24 brd 172.30.5.255 scope global dynamic noprefixroute eth0 valid_lft 2750sec preferred_lft 2750sec inet6 fe80::fa:d8ff:fe6a:ae1/64 scope link noprefixroute valid_lft forever preferred_lft forever
Looping in @clime - maybe he can help.
Huh. I'm not sure I see anything else to enable on the aws side. It should auto assign you one, but it seems like it's not. ;(
Perhaps @praiskup has some ideas?
For Copr we do this using ansible (works both for Fedora and RHEL9): https://pagure.io/fedora-infra/ansible/blob/4a01fb6f3e1da53bc8e535ba8c38e1b195a50975/f/roles/copr/base/tasks/main.yml#_110-126
And this when we migrate ipv6 from one box to another: https://github.com/fedora-copr/ansible-fedora-copr/blob/ddcf0719ef5f9ae30b0b7f7561828ee89b4aeb0b/tasks/reassign-ipv6.yml#L1-L12
Actually ~hour ago I enabled ipv6 for copr-pulp-dev, nagios green finally: https://nagios-external.fedoraproject.org/nagios/cgi-bin//status.cgi?hostgroup=copr_all_instances_aws&style=overview
Any news here @oliver ?
I can't find anything else off hand to change on our end... did you get it working? Is there anything else we can try?
@kevin Still trying to figure out what's going on. That said, it's a Fedora 37 system and I feel like I should be updating to a supported version before actually further trying to hunt down the problem. I hope the upgrade goes smooth - if not, you'll be hearing from me :-)
Alright. The update to F38 was straight forward and smooth, but didn't change anything in regards to IPv6. I'd say let's postpone this, until I got a bit more time to debug. It's really strange.
Metadata Update from @oliver: - Issue close_status updated to: Will Not/Can Not fix - Issue status updated to: Closed (was: Open)
ok. My next suggestion would be to deploy another instance and migrate to it... in case there's something weird with the existing one...
but we can debug it when you have more time. ;)
My apologies - the last weeks have been a roller coaster... I have not been able to get this running on the existing instance, so maybe a new instance is the way to go. I guess we can keep the external IP, right? Anyway, I'd like to postpone this action for a bit, until I have my head free again to catch up on these topics - currently, it's kinda impossible :-(
Yeah, I know how it goes for sure. ;(
Please just re-open or file a new ticket when you are ready and if we can do anything to help out.
Thanks for your assistance with this Kevin! I'll for sure reopen or open up a new request, once I'm ready.
Hi,
I think this issue is still valid. Pagure users libravatar.org by default, making deployed instances not completely operational in IPv6-only environments.
Agree with you, it's of course still relevant, but I'm currently drowning in work and only thinking about migrating the service from one host to another let's me panic and sweat :-)
Log in to comment on this ticket.