OVB overcloud deploy fails sporadically with "not enough free physical ports" error
Bug #1879472 reported by
Sagi (Sergey) Shnaidman
This bug affects 1 person
Affects | Status | Importance | Assigned to | Milestone | |
---|---|---|---|---|---|
tripleo |
Fix Released
|
Critical
|
Unassigned |
Bug Description
There are cases when OVB overcloud deploy fails on vexxhost with error like:
metalsmith.
It happens on vexxhost cloud only, about 10 times a day. Although we run promotions on vexxhost cloud only and it affect them.
Changed in tripleo: | |
milestone: | ussuri-rc3 → victoria-1 |
Changed in tripleo: | |
milestone: | victoria-1 → victoria-3 |
Changed in tripleo: | |
milestone: | victoria-3 → wallaby-1 |
Changed in tripleo: | |
milestone: | wallaby-1 → wallaby-2 |
Changed in tripleo: | |
milestone: | wallaby-2 → wallaby-3 |
Changed in tripleo: | |
status: | Triaged → Fix Released |
To post a comment you must log in.
I have been following the life cycle of one of the ports mentioned in the log (ec3ddedf- 3255-4f2f- ae8e-dd5b407322 a7) across all different logs: http:// paste.openstack .org/show/ 793977/
If I understood it correctly, the error is received by the undercloud log, so it is sent by the undercloud neutron server. This discards any issue with available ports in the Vexxhost tenant, but we still need to find out why it is only happening on Vexxhost.
We need someone from the Ironic team to have a look at the logs (and potentially check the OVB configuration at Vexxhost) to move forward.