Created attachment 1186053 [details] error logs where the error found Description of problem: Overcloud update failed with error Failed to detach interface Version-Release number of selected component (if applicable): openstack-heat-api-6.0.0-8.el7ost.noarch openstack-heat-api-cfn-6.0.0-8.el7ost.noarch openstack-tripleo-heat-templates-liberty-2.0.0-24.el7ost.noarch openstack-tripleo-heat-templates-kilo-0.8.14-16.el7ost.noarch How reproducible: Steps to Reproduce: 1.deploy rhos8 2.launch an instance from network 3.update the undercloud 4. update the overcloud Actual results: overcloud_update failed Expected results: Upgrade success Additional info: based on Deployment rhos8 overcloud openstack overcloud deploy --templates ~/templates/my-overcloud --control-scale 3 --compute-scale 1 --ntp-server clock.redhat.com --libvirt-type qemu -e ~/templates/my-overcloud/environments/network-isolation.yaml -e ~/templates/network-environment.yaml -e ~/templates/firstboot-environment.yaml update spawn instancetnet=$(neutron net-list | grep tenant-net | awk '{print $2};') nova boot --image cirros --flavor m1.tiny firstInstance --nic net-id=$tnet openstack overcloud deploy --templates ~/templates/my-overcloud --control-scale 3 --compute-scale 1 --ntp-server clock.redhat.com --libvirt-type qemu -e ~/templates/my-overcloud/environments/network-isolation.yaml -e ~/templates/network-environment.yaml -e /usr/share/openstack-tripleo-heat-templates/environments/major-upgrade-aodh.yaml --force-postconfig
According to the attached log, multiple compute and controller nodes are getting the detach error. This suggests that some change is causing port resources to be replaced. Chances are this is because other network resources are being replaced unexpectedly - this should be the first thing to check. Attaching an event list for the stack update should show what network resources are being replaced: heat event-list --nested-depth 3 --format log overcloud If those replacements are unexpected, then you'll need to figure out what properties are changing which cause replacement. If those replacements are expected and these nodes really do need new ports attached to them, then you'll need to diagnose nova for the cause of the detach failures. I'll have a look in upstream logstash to see if https://bugs.launchpad.net/heat/+bug/1585858 might be related
I'm not seeing any detach errors in upstream gate jobs
In an IRC conversation, it was brought up that there was a mistake in the process (the templates were copied to $HOME but not updated with the newer templates). Given that others have not seen this issue so far, we're going to close the bug. If it reproduces, please reopen the bug.