This will unify all the handlers to use first uppercase letter for
ansible-lint to stop complaining.
I went through all `notify:` occurrences and fixed them by running
```
set TEXT "text_to_replace"; set REPLACEMENT "replacement_text"; git grep
-rlz "$TEXT" . | xargs -0 sed -i "s/$TEXT/$REPLACEMENT/g"
```
Then I went through all the changes and removed the ones that wasn't
expected to be changed.
Fixes https://pagure.io/fedora-infrastructure/issue/12391
Signed-off-by: Michal Konecny <mkonecny@redhat.com>
In various roles I maintain I use `python3 -m pip install` to
directly install a Python project (usually a fedora-messaging
consumer), to avoid the pointless bureaucracy of packaging them.
The roles install all the deps of these projects as packages
first, so pip doesn't have to install any deps, it only installs
the project itself. Well...that's the idea. It's possible for
this to go wrong (say I forget to update the roles when adding
a dep to the project), and in that case I think we'd rather have
things blow up (so I know something's wrong) than have pip
silently install some random upstream wheel system-wide to make
it work. The intent is that all the deps still come from proper
Fedora packages, only these projects themselves get installed
directly.
Signed-off-by: Adam Williamson <awilliam@redhat.com>
nirik and I went around and around a bit today and ended up back
where we started, but with a clearer understanding of where that
this. This explains it a bit better, and makes what's actually
going on in various places clearer with the use of appropriate
shared variables. This should not actually *change* anything at
all when deployed.
Signed-off-by: Adam Williamson <awilliam@redhat.com>
In openqa/dispatcher, relvalconsumer and check-compose roles, we
install Python libraries from git checkouts (these are things we
don't really want to package as they change too much). This
enhances those roles so that we check whether pip considers the
libraries to be installed, and install them if it doesn't. The
purpose is to catch when the Python version rolls over on system
upgrade, and reinstall the libraries in that case - I got bitten
by this when upgrading to F32, I forgot to reinstall these libs
for Python 3.8, and it broke things for a couple of days before
I noticed and fixed it manually...
Signed-off-by: Adam Williamson <awilliam@redhat.com>
I changed how check-compose upstream does email recipient config
to try and fix the 'get Atomic-related emails to Atomic people'
problem again after Fedora-Atomic composes went away. This is
an attempt to adjust the play to populate the config file for
that change. Let's see what blows up!
Signed-off-by: Adam Williamson <awilliam@redhat.com>
This should hopefully avoid an awkward problem I noticed with
'python3 setup.py install' dumping replacements in /usr/local
for packaged scripts (e.g. fedmsg-logger)...
Signed-off-by: Adam Williamson <awilliam@redhat.com>
This includes some tweaks to the core fedmsg roles to allow a
'generic' way of indicating that a box should use fedmsg-hub-3
not fedmsg-hub, and make the restart notification work for that.
Signed-off-by: Adam Williamson <awilliam@redhat.com>
The special configuration for check-compose mails for two-week
Atomic nightly composes was broken due to fedfind changes. We
need to tweak this template a bit as part of fixing it up.
Signed-off-by: Adam Williamson <awilliam@redhat.com>
Update the check-compose config file to use the variant config
for the 'Atomic' milestone as well as the 'Postrelease'
milestone - this should DTRT for the new Pungi 4 two-week
Atomic composes, which have 'Atomic' as their milestone, sending
the report to the correct addresses.
I'm switching the misc. QA fedmsg consumers over to using fedmsg-
hub, due to https://github.com/fedora-infra/fedmsg/issues/365 .
So we need to adjust how we install check-compose, install a
config file to enable the consumer, and also set up the fedmsg
base and hub roles on the openqa server boxes (which do the
check-compose job ATM).
