The issue we have seems to be related to the fact that this unit test mixes:
- old engine facade code from networking-bagpipe OVO code
- new engine facade code for IPAllocations
My current understanding is that:
- [2] resulted in the use of OVO for IPAllocations
- networking-bgpvpn then started to make use of OVO, resulting in IPAllocation in OVO objects construction (both old engine facade, so all good)
- when [1] merged to revert [2] because of bug 174829, IPAllocation went back to non-OVO, hence
using the new engine facade, while we still use OVO to query IPAllocations in networking-bagpipe OVO objects
The issue we have seems to be related to the fact that this unit test mixes:
- old engine facade code from networking-bagpipe OVO code
- new engine facade code for IPAllocations
This issue is strongly related to bug 1744829 .
My current understanding is that:
- [2] resulted in the use of OVO for IPAllocations
- networking-bgpvpn then started to make use of OVO, resulting in IPAllocation in OVO objects construction (both old engine facade, so all good)
- when [1] merged to revert [2] because of bug 174829, IPAllocation went back to non-OVO, hence
using the new engine facade, while we still use OVO to query IPAllocations in networking-bagpipe OVO objects
[1] https:/ /review. openstack. org/#/c/ 536913 /review. openstack. org/#/c/ 407868
[2] https:/