macvtap agent: wrong vif details when instance is launched on vlan and flat network

Bug #1745159 reported by Andreas Scheuring on 2018-01-24
6
This bug affects 1 person
Affects Status Importance Assigned to Milestone
neutron
High
Unassigned

Bug Description

It's bug in the macvtap mechanism driver.

If multiple network types are used with the macvtap driver, e.g. flat and vlan, the vif_details can contain an invalid vif_detail attribute 'vlan'. This leads to nova configuring the macvtap with a vlan, even though flat should be used!

The reason for this is, that the vif_details are supposed to be the same for all ports managed by a mechanism driver. Due to that they are stored as instance variables. However they are not with macvtap. In the vlan case, the vlan, which is stored in the vif_details, of course depends on the network. It might be a different for port A than for port B. The problem is, that the driver is always updating the global vif_dict, but instead of it should operate on a local copy. Now if first a vlan port is processed and after that a flat port, the vlan vif_detail is still stored in the mechanism drivers global vif_details. A flat port becomes a vlan port.

Fix proposed to branch: master
Review: https://review.openstack.org/537407

Changed in neutron:
assignee: nobody → Andreas Scheuring (andreas-scheuring)
status: New → In Progress
Akihiro Motoki (amotoki) on 2018-01-30
tags: added: macvtap
Akihiro Motoki (amotoki) wrote :

IIUC, it prevents using two networks with different network types with macvtap driver.

Changed in neutron:
importance: Undecided → High

exactly!

This bug has had a related patch abandoned and has been automatically un-assigned due to inactivity. Please re-assign yourself if you are continuing work or adjust the state as appropriate if it is no longer valid.

Changed in neutron:
assignee: Andreas Scheuring (andreas-scheuring) → nobody
status: In Progress → New
tags: added: timeout-abandon

Change abandoned by Slawek Kaplonski (<email address hidden>) on branch: master
Review: https://review.openstack.org/537407
Reason: This review is > 4 weeks without comment, and failed Jenkins the last time it was checked. We are abandoning this for now. Feel free to reactivate the review by pressing the restore button and leaving a 'recheck' comment to get fresh test results.

To post a comment you must log in.
This report contains Public information  Edit
Everyone can see this information.

Other bug subscribers