On 11/03/19 18:05 +0100, Jan Pokorný wrote:
On 11/03/19 15:01 +0100, Jan Pokorný wrote:
> On 11/03/19 12:31 +0100, Vít Ondruch wrote:
>> Can somebody please enlighten me, how to update Rawhide after branching
>> and not using --nogpgcheck?
>
> Good question; have observed this over several (if not all since
> to-be-f26 based Rawhide) jumps like that, and always have solved this
> inconvenience by hand and moved on.
>
> GPG signatures related must-do consequence of branching for those
> using mock with the branched + Rawhide targets is also to update
> mock-core-configs, which is currently also not very straightforward
> and could perhaps be handled more gracefully:
>
https://bugzilla.redhat.com/show_bug.cgi?id=1686869
Actually, fedora:rawhide container from Docker Hub (in CI scenario
here) is affected with this problem as well (cannot update with the
new-key-signed f30/f31 packages because of this, making the CI
procedure fail[*]).
Should not at least the "official container images" follow the
branching point automatically and promptly to prevent GPG signature
imposed disruptions ASAP?
Not sure if this is thanks to the images indeed being
periodically updated (image's sha256 changed from
1c2065c021750bfa8b1530d2cce08dffda03f0fbe5fba4f21b29fc8e909954a9 to
eff97b05aaa6ef5b3444fc95cc8b340266ea63474dca3f557f07ccfd70f1fee1
between failed vs. successful run), but everything is back to normal
now, and this symptom of the failed "dnf update" with fedora:rawhide
went away, too:
Importing GPG key 0xCFC659B9:
Userid : "Fedora (30) <fedora-30-primary(a)fedoraproject.org>"
Fingerprint: F1D8 EC98 F241 AAF2 0DF6 9420 EF3C 111F CFC6 59B9
From : /etc/pki/rpm-gpg/RPM-GPG-KEY-fedora-30-x86_64
Key imported successfully
Import of key(s) didn't help, wrong key(s)?
All in all, Would be preferred if this was glitch-free all the time.
--
Jan (Poki)