Exception on subcloud admin address pool delete
Affects | Status | Importance | Assigned to | Milestone | |
---|---|---|---|---|---|
StarlingX |
Fix Released
|
Low
|
Lucas Ratusznei Fonseca |
Bug Description
Brief Description
-----------------
System AIO-SX (virtualbox) subcloud initially bootstrapped with the admin network.
When an admin address pool is deleted (system addrpool-delete), the expected behaviour is that a subcloud falls back to using the management network.
An exception is occurring when performing this activity after the integration of the L3 firewall feature:
In the above scenerio, mgmt and cluster-host are still on the loopback interface (as the admin network is handling communication to/from the system controller). So the only network in firewall_networks is the oam network. Thus, referencing the network dict for mgmt will cause an exception.
This code is currently contained in get_host_config(). This is called any time the puppet hieradata needs to get updated. That is, whenever a runtime manifest needs to go (this case), or other activities like host lock/unlock.
Severity
--------
Minor: System/Feature is usable with minor issue
Reproducibility
---------------
100% reproducible
Related to recent code changes for: https:/ /storyboard. openstack. org/#!/ story/2010591
Fixed by: https:/ /review. opendev. org/c/starlingx /config/ +/889125