delete interface times out in check-tempest-dsvm-neutron

Bug #1291677 reported by Brant Knudson
26
This bug affects 4 people
Affects Status Importance Assigned to Milestone
tempest
Invalid
Undecided
Unassigned

Bug Description

This occurred during a tempest run on https://review.openstack.org/#/c/78459/ (in Keystone). check-tempest-dsvm-neutron failed with a single test failure: tempest.api.compute.v3.servers.test_attach_interfaces.AttachInterfacesV3Test.test_create_list_show_delete_interfaces

Looks like the test gets a token, starts booting an instance:

POST http://127.0.0.1:5000/v2.0/tokens - Status: 200
POST http://127.0.0.1:8774/v3/servers - Status: 202

... this is all probably expected... eventually it does a DELETE:

2014-03-12 20:59:32,797
Request: DELETE http://127.0.0.1:8774/v3/servers/f4433c4f-4d27-492d-8794-77674a634c3f/os-attach-interfaces/f145579d-aaa5-4d44-941c-9856500f65b5
Response Status: 202

Then it starts requesting status... and eventually it gives up:

2014-03-12 21:02:48,016
GET http://127.0.0.1:8774/v3/servers/f4433c4f-4d27-492d-8794-77674a634c3f/os-attach-interfaces
Response Status: 200
Response Body: ... "port_state": "ACTIVE", ... "port_state": "ACTIVE" ... "port_state": "ACTIVE"

The test failed in _test_delete_interface:

File "tempest/api/compute/v3/servers/test_attach_interfaces.py", line 127, in test_create_list_show_delete_interfaces
File "tempest/api/compute/v3/servers/test_attach_interfaces.py", line 96, in _test_delete_interface
Details: Failed to delete interface within the required time: 196 sec.

The timings at the end show the slowpoke:

tempest.api.compute.v3.servers.test_attach_interfaces.AttachInterfacesV3Test.test_create_list_show_delete_interfaces[gate,smoke] 218.999

------------

Neutron's last entry for the instance in q-svc.txt is

[12/Mar/2014 21:02:52] "GET /v2.0/ports.json?device_id=f4433c4f-4d27-492d-8794-77674a634c3f HTTP/1.1" 200 1912 0.027403

I grepped through the q-svc.txt log file and I don't see anyplace where the DELETE is forwarded on.

--------

Did the test not wait long enough for the operation to complete? Was the shutdown request ignored?

Revision history for this message
Matt Riedemann (mriedem) wrote :
Aaron Rosen (arosen)
tags: added: network
Aaron Rosen (arosen)
Changed in nova:
assignee: nobody → Aaron Rosen (arosen)
importance: Undecided → High
Aaron Rosen (arosen)
Changed in nova:
assignee: Aaron Rosen (arosen) → nobody
importance: High → Undecided
Revision history for this message
Salvatore Orlando (salvatore-orlando) wrote :

This has exactly the same signature as bug 1291992.
However this has an e-r query which even if different from the one prosed (but not approved) for bug 1291992 has apparently a decent correlation.

1291992 will be therefore marked as duplicate (this was filed first after all!) and the assignee moved to this bug.

Changed in neutron:
importance: Undecided → High
tags: added: neutron-rc-potential
Changed in neutron:
assignee: nobody → Aaron Rosen (arosen)
Revision history for this message
Eugene Nikanorov (enikanorov) wrote :

Logs are not available and issue has not been seen in the gate for a few months.
Hence marking bug as incomplete.

Changed in neutron:
status: New → Incomplete
Sean Dague (sdague)
no longer affects: nova
no longer affects: neutron
Changed in tempest:
status: New → Invalid
To post a comment you must log in.
This report contains Public information  
Everyone can see this information.

Duplicates of this bug

Other bug subscribers

Remote bug watches

Bug watches keep track of this bug in other bug trackers.