Avoid dhcp-all-interfaces durin update/upgrade
Affects | Status | Importance | Assigned to | Milestone | |
---|---|---|---|---|---|
tripleo |
Triaged
|
Medium
|
Unassigned |
Bug Description
When there is a failure in os-net-config, THT has a fallback mechanism to DHCP on all interfaces to ensure that connection is not lost. This works fine for a fresh deployment. But during an update or upgrade, when the DHCP all interfaces is triggered, because of temproray failure (like ping metada in queens), it will lead to openstack service failures.
Instead, it is apt to either DHCP only the control plane network or default network interfaces. OR avoid the dhcp all interfaces during an update/upgarde.
Additionally, analyze if failure recovery can be done in the os-net-config itlsef with intelligence of identifying which interface to initiate DHCP.
Detailed discussion on this BZ https:/
Changed in tripleo: | |
milestone: | none → ussuri-2 |
importance: | Undecided → Medium |
status: | New → Triaged |
Changed in tripleo: | |
milestone: | ussuri-2 → ussuri-3 |
Changed in tripleo: | |
milestone: | ussuri-3 → ussuri-rc3 |
Changed in tripleo: | |
milestone: | ussuri-rc3 → victoria-1 |
Changed in tripleo: | |
milestone: | victoria-1 → victoria-3 |