- the task container needs to wait longer for migrations to complete for fresh installs before starting services
- otherwise, services start prematurely and clutter the logs with errors because migrations are mid-flight
* Our tests could consistently get awx jobs into a deadlocked state
whenever the parallelism was high. Even podman ps would hang when the
system was in this state. We don't know exactly where in runc the bug is
but the deadlocks stopped happening when we changed the OCI runtime
environment to crun.
Jobs unable to start because podman trying to use systemd cgroup manager. See error below :
```
WARN[0000] Failed to add conmon to systemd sandbox cgroup: dial unix /run/systemd/private: connect: no such file or directory
Error: OCI runtime error: systemd cgroup flag passed, but systemd support for managing cgroups is not available
```
* According to this PR https://github.com/containers/podman/pull/7009, podman switch references from libpod.conf to containers.conf.
* According to containers.conf man (https://github.com/containers/common/blob/main/docs/containers.conf.5.md), configuration file is a TOML file but engine section declaration is missing.
Force fully qualified image names
If we try and pull an unqualified image name, jobs hang on a podman
prompt.
I set the permissions as 644 because thats what worked for me because rootless podman needs to be able to read the file, but maybe there is another way to achieve that
Reviewed-by: Christian Adams <rooftopcellist@gmail.com>
Isolated removal
SUMMARY
Removal of the isolated nodes feature.
ISSUE TYPE
Feature Pull Request
COMPONENT NAME
API
AWX VERSION
Reviewed-by: Alan Rominger <arominge@redhat.com>
Reviewed-by: Jeff Bradberry <None>
Reviewed-by: Elyézer Rezende <None>
Reviewed-by: Bianca Henderson <beeankha@gmail.com>
Remove invocation of Ansible from launch script
This was missed in the initial EE PR. Ansible is no longer installed inside of the web & task containers, causing this to show up in the container logs:
/usr/bin/launch_awx.sh: line 18: ansible: command not found
/usr/bin/launch_awx.sh: line 19: ansible: command not found
Reviewed-by: Ryan Petrello <None>
This was missed in the initial EE PR. Ansible is no longer installed inside of
the web & task containers, causing this to show up in the container logs:
```
/usr/bin/launch_awx.sh: line 18: ansible: command not found
/usr/bin/launch_awx.sh: line 19: ansible: command not found
```
- remove requirements_ansible logic from the update script
- removed the need for py2-specific system dependencies
- update to the latest pip-tools and move to the new long format
(https://github.com/jazzband/pip-tools/pull/1237)
- fixed a few busted references to receptorctl @ devel
- removes local_docker installer and points community users to our development environment (make docker-compose)
- provides a migration path from Local Docker Compose installations --> the dev environment
- the dev env can now be configured to use an external database
- consolidated the Local Docker and dev env docker-compose.yml files into one template file, used by the dockerfile role
- added a 'sources' role to template out config files
- the postgres data dir is no longer a bind-mount, it is a docker volume
- the redis socket is not longer a bind-mount, it is a docker volume
- the local_settings.py.docker-compose file no longer needs to be copied over in the dev env
- Create tmp rsyslog.conf in rsyslog volume to avoid cross-linking. Previously, the tmp code-generated rsyslog.conf was being written to /tmp (by default). As a result, we were attempting to shutil.move() across volumes.
- move k8s image build and push roles under tools/ansible
- See tools/docker-compose/README.md for usage of these changes
