As of 2025-12-07 00:00 JST(=UTC+9h), fedora git repository seems very unstable or unreachable.
For example, $ git pull takes very long time, or sometimes returns:
$ git pull
Error during lookup request: status: 503 fatal: Could not read from remote repository. Please make sure you have the correct access rights and the repository exists.
And this seems to be affecting bodhi build using rpkg, for example: https://copr.fedorainfracloud.org/coprs/mtasaka/rubygems-40-all-02/build/9880709/
rpkg
builder-live.log.gz says:
cmd: ['git', 'clone', 'https://src.fedoraproject.org/rpms/rubygem-aruba.git', '/var/lib/copr-rpmbuild/workspace/workdir-8hg6n96r/rubygem-aruba', '--depth', '500', '--no-single-branch', '--recursive'] cwd: . rc: 128 stdout: stderr: Cloning into '/var/lib/copr-rpmbuild/workspace/workdir-8hg6n96r/rubygem-aruba'... error: RPC failed; HTTP 503 curl 22 The requested URL returned error: 503 fatal: the remote end hung up unexpectedly Cloning into '/var/lib/copr-rpmbuild/workspace/workdir-8hg6n96r/rubygem-aruba'... error: RPC failed; HTTP 503 curl 22 The requested URL returned error: 503 fatal: the remote end hung up unexpectedly
Would you investigate the current git server status? Thank you in advance.
I also encountered the similar instability issue, when git from command line or on fedora package sources website.
I'm also running into this issue:
https://koji.fedoraproject.org/koji/taskinfo?taskID=139758169 => https://kojipkgs.fedoraproject.org//work/tasks/8217/139758217/checkout.log =>
$ git clone -n https://src.fedoraproject.org/rpms/python-inema.git /var/lib/mock/f42-build-64258467-6623784/root/chroot_tmpdir/scmroot/python-inema Cloning into '/var/lib/mock/f42-build-64258467-6623784/root/chroot_tmpdir/scmroot/python-inema'... fatal: unable to access 'https://src.fedoraproject.org/rpms/python-inema.git/': The requested URL returned error: 503
https://src.fedoraproject.org/rpms/python-inema.git/ => Firefox get on my workstation currently yields:
Error 503 Backend fetch failed Backend fetch failed Guru Meditation: XID: 1477804044 Varnish cache server
https://www.fedorastatus.org/ claims that everything is fine, though: =>
All Systems Operational
Visiting the front page https://src.fedoraproject.org/ in a browser takes very long and ends with an unavailable page, right now:
Sorry! This service is currently unavailable.
And https://www.fedorastatus.org/ still reports that everything is fine. Either that service doesn't health check src.fedoraproject.org or that check isn't very thorough.
Yes, it's scrapers. ;(
And https://www.fedorastatus.org/ still reports that everything is fine.
Yes, it's manually updated when we know about a problem. Luckily in my sleep I didn't know about this. ;( (nightmare?)
will start looking now.
Metadata Update from @kevin: - Issue assigned to kevin - Issue priority set to: Waiting on Assignee (was: Needs Review) - Issue tagged with: high-gain, high-trouble
ok, the problem seems to have been that scrapers were (aggressively as usual) crawling the kernel package. They were hitting /blame/ and /history/ endpoints on every commit page they fetched, then recursively doing so on those and so on. ;(
I adjusted a bunch of things (more memory for pkgs01 backend, more apache threads/workers, caching of ttf files), but in the end the thing that got things back to normal was just blocking the /blame/ and /history/ endpoints. So, the web interface will no longer offer those (at least for now). If you need to do those things, you can git clone locally and do them easily.
So, things should be back to normal now and hopefully will stay that way.
Metadata Update from @kevin: - Issue close_status updated to: Fixed - Issue status updated to: Closed (was: Open)