2024-10-19T09:08:11  *** teepee_ is now known as teepee
2024-10-19T13:12:01  *** teepee_ is now known as teepee
2024-10-19T13:59:22  *** teepee_ is now known as teepee
2024-10-19T17:34:31  *** teepee_ is now known as teepee
2024-10-19T17:48:36  <acidsys> bmwiedemann: could you check https://progress.opensuse.org/issues/168499 please
2024-10-19T19:14:03  <matrix-o-o> <SFaulken> LCP: is there anything I can do, to help clean out the recents on paste.o.o?
2024-10-19T19:15:10  <matrix-o-o> <LCP> I didn't have the time to give you perms to do it, sorry
2024-10-19T19:15:22  <matrix-o-o> <SFaulken> no worries.
2024-10-19T20:18:02  <acidsys> any ideas https://github.com/openSUSE/get-o-o/actions/runs/11420435256/job/31776187919?pr=226 => "SocketError: Failed to open TCP connection to mirror.rackspace.com:443", is that thing supposed to have access to that mirror
2024-10-19T20:21:42  <cboltz> I'd guess it checks download.o.o for the image sizes, and gets redirected to random mirrors
2024-10-19T20:22:02  <acidsys> in that case, it is supposed to have access to arbitrary mirrors download.o.o redirects to
2024-10-19T20:22:07  <acidsys> s/it is/is it/
2024-10-19T20:24:44  <cboltz> depends on the redirect target it gets from download.o.o, so yes, it might need access to random mirrors
2024-10-19T20:26:42  <acidsys> so who does that? it's some github managed server if I understand correctly
2024-10-19T20:27:15  <cboltz> have a look at the get-o-o Rakefile:
2024-10-19T20:27:27  <cboltz>             conn = ::Faraday.new(url: URI("https://download.opensuse.org#{k['primary_link']}")) do |r|
2024-10-19T20:27:32  <cboltz>               r.use ::FaradayMiddleware::FollowRedirects, limit: 5
2024-10-19T20:28:43  <cboltz> also note that download.o.o might not do redirects when we build get.o.o in our infrastructure because it recognizes our source IP as internal
2024-10-19T20:28:58  <cboltz> while github servers obviously are not internal
2024-10-19T20:37:55  <acidsys> sorry, but I don't understand how that helps me make the PR pipeline green ;-)
2024-10-19T20:38:08  <acidsys> in previous PRs it seems to have worked
2024-10-19T20:39:19  <cboltz> restarting the pipeline might help - either the mirror recovered, or you get redirected to another mirror
2024-10-19T20:40:40  <acidsys> happens again https://github.com/openSUSE/get-o-o/actions/runs/11420435367/job/31776483273
2024-10-19T20:40:59  <acidsys> the mirror seems rather fine to me
2024-10-19T20:42:16  <acidsys> according to https://docs.github.com/en/actions/using-github-hosted-runners/connecting-to-a-private-network/about-private-networking-with-github-hosted-runners#about-github-hosted-runners-networking it should have internet access
2024-10-19T20:42:55  <cboltz> right
2024-10-19T20:48:29  <cboltz> after searching for the error message " SocketError: getaddrinfo: Try again", we are at an old saying:   it's always DNS   ;-)
2024-10-19T20:49:36  <cboltz> the DNS server used by github might be overloaded etc. - but not completely broken because the lookup done for download.o.o seems to have worked
2024-10-19T20:52:32  <acidsys> fun
2024-10-19T20:52:59  <acidsys> ok to merge regardless? is just one word in the body..
2024-10-19T20:54:28  <cboltz> yes, I'll take the risk ;-)
2024-10-19T20:54:51  <acidsys> even better, not in my name then ;)
2024-10-19T21:03:18  <acidsys> bmwiedemann: fyi I also asked darix regarding stage