Unable to add new compute nodes after OpenStack 13 to 16.x upgrade due to NIC naming
Issue
- After an upgrade from OpenStack 13 to 16.x, compute nodes use older RHEL7 NIC names as a result of the LEAPP upgrade.
- New nodes with identical hardware will use default RHEL8 names for NICs .
- For example, an upgraded compute node will use the RHEL7 NIC name such as
enp55s1f0
, but a new, identical node will use the nameenp55s1f8
for the same NIC. - The OpenStack deployment configuration ("templates") use explicit interface names in the NIC configs.
- If OpenStack templates are updated with the NIC names for the new nodes, then existing compute nodes will fail network configuration, resulting in nodes that will not boot.
Environment
- Red Hat OpenStack 16.x that has been upgraded from Red Hat OpenStack 13
- Red Hat Enterprise Linux 8 compute nodes that have been upgraded from Red Hat Enterprise Linux 7 using LEAPP
Subscriber exclusive content
A Red Hat subscription provides unlimited access to our knowledgebase, tools, and much more.