Tempest test case tempest.api.volume.test_volumes_actions.VolumesV2ActionsTest fails with NetAPP + NFS setup in Juno based

[novelkumar]

Setup:

June based installation with NetAPP NFS as storage backend. Glance images are stored in Swift.

Issue:

Tempest test case tempest.api.volume.test_volumes_actions.VolumesV2ActionsTest fails with timeout error as shown below:

Traceback (most recent call last):
  File "/opt/stack/tempest/tempest/test.py", line 112, in wrapper
    return f(self, *func_args, **func_kwargs)
  File "/opt/stack/tempest/tempest/api/volume/test_volumes_actions.py", line 104, in test_volume_upload
    self.image_client.wait_for_image_status(image_id, 'active')
  File "/opt/stack/tempest/tempest/services/image/v1/json/image_client.py", line 316, in wait_for_image_status
    raise exceptions.TimeoutException(message)
TimeoutException: Request timed out
Details: (VolumesV2ActionsTest:test_volume_upload) Time Limit Exceeded! (300s)while waiting for active, but we got saving.

It seems the portion of cinder volume create and uploads to glance image is failing. glance image-list shows the newly created image in 'saving' status
for quite a long time [more than 30 minutes]. But once the tempest test fails, the new image is seems as active in glance.

A manual testing [volume creation and uploading the volume to glance] is pretty fast compared to tempest testing.

Is there a timeout value which i can increase for tempest testing? Or is this an issue with tempest test method?

The cinder-volume log shows the following snippet for the volume UUID 40fdecd0-3de3-43b7-8a16-01a9f4b1c50e:

[novelkumar (novel-kumar)]

The cinder-volume log shows the following snippet for the volume UUID 40fdecd0-3de3-43b7-8a16-01a9f4b1c50e:

[novelkumar (novel-kumar)]

The trace occurs after the following warning and error in cinder-volume log:

2015-08-06 07:29:20.326 1508 WARNING requests.packages.urllib3.connectionpool [req-5eea8cd6-cde0-4949-8f9c-ff46427863fb 315fb53eb8ba4671a81a9d5815e8a504 4139981c695949af8b86b3e36e66dcff - - -] Connection pool is full, discarding connection: 10.1.38.18
2015-08-06 07:29:20.328 1508 ERROR cinder.volume.manager [req-5eea8cd6-cde0-4949-8f9c-ff46427863fb 315fb53eb8ba4671a81a9d5815e8a504 4139981c695949af8b86b3e36e66dcff - - -] Error occurred while uploading volume 40fdecd0-3de3-43b7-8a16-01a9f4b1c50e to image 2118ec39-bedf-4075-829d-2c18b7beeb51.

[Jordan Pittier (jordan-pittier)]

Neither Juno nor the NFS driver by NetApp are supported here nowadays. Please contact your vendor if you need assistance.