We just updated the scheduler code to handle job re-trigger
requests, we need to configure the listener to listen for the
appropriate messages.
Signed-off-by: Adam Williamson <awilliam@redhat.com>
We defined all these in an inventory variables, but also listed
the same keys explicitly in the queue configs in the group
playbook. Let's not do that.
Signed-off-by: Adam Williamson <awilliam@redhat.com>
Already in effect on prod & stg, formalize these configuration tweaks
that limit overload, improve cgroup monitoring, and build tolerance to
transient fileserver problems.
Turns out the current version of linux-system-roles network doesn't like
setting ipv6 no and setting search domains since NetworkManager wants to
set them for both ipv4 and ipv6. There's some upstream fix, but not in
our current version.
Signed-off-by: Kevin Fenzi <kevin@scrye.com>
This was done using yq (
https://mikefarah.gitbook.io/yq/operators/sort-keys )
Doing things this way makes it much easier to see if a variable is set
in a file or if two hosts differ in what variables they set. Hopefully
we can keep things sorted moving forward.
Basically this means just sort a-z anything you add to any host or group
vaiable and it will be in the right place.
Additionally, this enforces 'normal' intent rules for all the variable
files which we should also try and obey. 2 spaces for first level, 3 for
next, etc. When in doubt you can run yq on it.
This should cause NO actual vairable changes, it's all just readability
fixing for humans, ansible parses it exactly the same.
Signed-off-by: Kevin Fenzi <kevin@scrye.com>
Currently we still see some errors with scanning the primary centos
stream 9 mirror.
The current assumption is that because of '--delay-updates' rsync
updates the primary mirror timestamps even if the actual files are moved
to the final location much later. This breaks our assumption during
scanning that we only have to check files in directories with changed
timestamps.
Currently we might scan the primary mirror and see updated timestamps
but the files are not yet moved to the final location because of
'--delay-updates'. After our scan, rsync finishes on the primary mirror
and the files are moved to the final location, but the timestamps of the
directories do not change. That results in us not detecting new files.
This change checks the timestamp of the COMPOSE_ID file and only runs
the scan if the timestamp is newer than during the last scan. The hope
is that the timestamp of COMPOSE_ID is only updated once the actual new
file is replacing the old file.
In addition to hopefully fixing the primary mirror scanning this should
also reduce the I/O load on the primary mirror because we do not run a
full rsync directory listing if COMPOSE_ID has not changed.
Signed-off-by: Adrian Reber <adrian@lisas.de>
I'm not sure what is causing some users to get /home/user instead of
/home/fedora/user. The ipa default is /home/fedora. Noggin doesn't
appear to override anything.
In any case fedorapeople is where this is really important as we have
web config and the like expecting /home/fedora/user. So, lets override
in sssd to make sure it uses the right place.
I'm also going to manually fix those users that are currently wrong.
See infra 10315
Signed-off-by: Kevin Fenzi <kevin@scrye.com>
