After reading some source code of charmed-openstack-tester and zaza-
openstack-tests, I finally set up a debug env by setting keep-
workspace=true, which prevents the tempest workspace from being deleted
after running charmed-openstack-tester. This allows me to quickly debug
a single failed test by running the following similar tempest command.

1, tempest run --workspace zaza --regex
'octavia_tempest_plugin.tests.scenario.v2.test_traffic_ops.TrafficOperationsScenarioTest'

I got the following two kinds of logs

#the log from ~/.tempest/zaza/tempest.log
2025-06-13 03:29:56.706 310372 INFO tempest.lib.common.ssh [-] Creating ssh 
connection to '10.149.142.113:22' as 'cirros' with public key authentication
2025-06-13 03:29:56.708 310372 INFO paramiko.transport [-] Connected (version 
2.0, client dropbear_2020.81)
2025-06-13 03:29:56.808 310372 INFO paramiko.transport [-] Authentication 
(publickey) successful!
2025-06-13 03:29:56.809 310372 INFO tempest.lib.common.ssh [-] ssh connection 
to [email protected] successfully created
2025-06-13 03:29:56.816 310372 INFO octavia_tempest_plugin.tests.validators [-] 
Validate URL got exception: HTTPConnectionPool(host='10.149.142.113', port=80): 
Max retries exceeded with url: / (Caused by 
NewConnectionError('<urllib3.connection.HTTPConnection object at 
0x7f961dec6860>: Failed to establish a new connection: [Errno 111] Connection 
refused')). Retrying
...
2025-06-13 03:44:54.182 310372 INFO octavia_tempest_plugin.tests.validators [-] 
Validate URL got exception: HTTPConnectionPool(host='10.149.142.113', port=80): 
Max retries exceeded wit
h url: / (Caused by NewConnectionError('<urllib3.connection.HTTPConnection 
object at 0x7f961dec70a0>: Failed to establish a new connection: [Errno 111] 
Connection refused')). Retrying.
2025-06-13 03:44:59.188 310372 INFO tempest.test [-] <class 
'tempest.lib.exceptions.TimeoutException'> raised in 
TrafficOperationsScenarioTest.setUpClass. Invoking tearDownClass.
                  
#the log from  /var/log/octavia/octavia-worker.log
...
                                                       |__Flow 
'octavia-create-loadbalancer-flow': 
octavia.common.exceptions.ComputeWaitTimeoutException: Waiting for compute id 
6246b204-53e3-4362-9351-1afdf99deb96 to go active timeout.
2025-06-13 02:15:47.862 198168 ERROR 
octavia.controller.worker.v2.controller_worker Traceback (most recent call 
last):
2025-06-13 02:15:47.862 198168 ERROR 
octavia.controller.worker.v2.controller_worker   File 
"/usr/lib/python3/dist-packages/taskflow/engines/action_engine/executor.py", 
line 53, in _execute_task
2025-06-13 02:15:47.862 198168 ERROR 
octavia.controller.worker.v2.controller_worker     result = 
task.execute(**arguments)
2025-06-13 02:15:47.862 198168 ERROR 
octavia.controller.worker.v2.controller_worker   File 
"/usr/lib/python3/dist-packages/octavia/controller/worker/v2/tasks/compute_tasks.py",
 line 302, in execute
2025-06-13 02:15:47.862 198168 ERROR 
octavia.controller.worker.v2.controller_worker     raise 
exceptions.ComputeWaitTimeoutException(id=compute_id)
2025-06-13 02:15:47.862 198168 ERROR 
octavia.controller.worker.v2.controller_worker 
octavia.common.exceptions.ComputeWaitTimeoutException: Waiting for compute id 
6246b204-53e3-4362-9351-1afdf99deb96 to go active timeout.

These logs show that TrafficOperationsScenarioTest is creating an
octavia backend, but timed out while verifying the service on port 80.
This is clearly unrelated to our metadata fix, as the backend VM is
accessible via SSH as '[email protected]' with public key
authentication, so this octavia_tempest_plugin test
failure(octavia_tempest_plugin.tests.scenario.v2.test_traffic_ops.TrafficOperationsScenarioTest)
is unrelated to metadata and metadata fix.

2, The remaining 3 heat_tempest_plugin tests are dynamically generated by 
gabbits, this is the output of running the following gabbits tests - 
https://paste.ubuntu.com/p/sknr3sgtBG/

tempest run --config-file etc/tempest.conf --regex
heat_tempest_plugin.tests.api.test_heat_api > heat_tempest.log 2>&1

These 3 failed gabbits tests just creates a stack instance of type
OS::Heat::TestResource for testing the heat api
(https://github.com/openstack/heat-tempest-
plugin/blob/master/heat_tempest_plugin/tests/api/gabbits/resources.yaml#L27),
it doesn't create a real VM, so it does not involve metadata.

3, Besides, We have also tested before that running charmed-openstack-
tester without this metadata fix result in these 4 failed tests as well.

4, And this morning, we received feedback from our customer, who
comfirmed: We tested the new neutron version 2:20.5.0-0ubuntu2.1 on one
of the compute hosts in a focal/yoga cluster and tested creation of
several VMs. This is working as expected.

Overall, I think this fix is fine. thank you.

** Tags removed: verification-needed-jammy verification-yoga-needed
** Tags added: verification-done-jammy verification-yoga-done

-- 
You received this bug notification because you are a member of Ubuntu
Bugs, which is subscribed to Ubuntu.
https://bugs.launchpad.net/bugs/2017748

Title:
  [SRU] OVN:  ovnmeta namespaces missing during scalability test causing
  DHCP issues

To manage notifications about this bug go to:
https://bugs.launchpad.net/cloud-archive/+bug/2017748/+subscriptions


-- 
ubuntu-bugs mailing list
[email protected]
https://lists.ubuntu.com/mailman/listinfo/ubuntu-bugs

Reply via email to