The magnum section is used to declare configuration specific information
used by magnum-tempest-plugin
The keys used are:
- nic_id, to indicate the external network
- image_id, to indicate the fedora-coreos image to be used
- flavor_id, the flavor id to use when creating clusters
- dns_nameserver, the upstream dns server IP.
- network_driver, the network driver to test (flannel).
- labels, to pass a custom (local) image registry.
- insecure_registry, to mark the custom image registry as http (instead
of https).
OpenStack Magnum relies on specific versions of Fedora CoreOS, this
patch addresses this maintaining a map of images per release according
to the upstream documentation[0]
The images are expected to be stored in the object store pointed out by
TEST_SWIFT_IP[1] environment variable in a container named 'magnum'. A
bash script to upload images can be found at
./zaza/openstack/charm_tests/magnum/upload_fedora_coreos_images.sh
[0] https://docs.openstack.org/magnum/latest/user/index.html#supported-versions
[1] https://github.com/openstack-charmers/zosci-config/pull/262
This test is unstable, increasing the logging from debug to info allows
to the CI runs to expose more information and understand what part of
the test is staying blocked and leading the CI job to time out.
Related-Bug: http://pad.lv/2003775
There is a race in the 408 test for rabbitmq where the config-change to
enable ssl causes a leader-settings-changed hook in the non-leader units
which results in a rabbitmq service restart. This can happen at exactly
the same time as the test attempts to establish a connection with the
that unit. This patch retries the connection attempt.
Note that this may only be a partial fix as it's possible that a restart
will happen just after the connection is made, which would then result
in a test failure.
Related-Bug: LP#2002156
There are scenarios where the config-changed hook can complete, yet the
service IP get configured many seconds after, because a relation-changed
hook execution needs to be triggered on the hacluster side of the
relation.
This change adds a retry to the check (10 times with a 2 seconds wait
time).
This issue was found at the gate https://review.opendev.org/c/openstack/charm-designate-bind/+/861417
The create_segment() function is often failing in the gate
due to being unable to establish a connection to the masakari
endpoint. This will allow some more time for the endpoint to
become available when this error occurs.
PyYAML<5.1 fails with the error below on Python 3.10
AttributeError: module 'collections' has no attribute 'Hashable'
The fix got available by this commit
9959328b41
nosetest doesn't support python 3.10 and over the years the python
community has been settling on pytest as test runner, this change
introduces pytest and adds py3.10 to the gh workflow testing matrix.
In some testing environment, `wait_for_application_states()` can
execute before juju starts actually performing changes on the
units, causing it to return immediately.
