Description of problem: When updating two ports on a switch concurrently, one fails because the other one holds lock. We get traceback. 2018-11-22 00:10:37.222 26 ERROR networking_ansible.ansible_networking [req-7435b1fa-dabb-42e3-82da-131d694dfd05 61467bdcf96a4e2b84bffd08b769f11d 7c5ad5bd6d9d4369a0a8a981b51695e1 - default default] Failed to plug in port 5b52a594-c654-4e60-b3f7-9a5a341d53eb on device: junos from network dd73584d-56e8-4d92-b7cd-1be30cb56e56 reason: ^M PLAY [Openstack networking-ansible playbook] *********************************** ^M TASK [openstack-ml2 : check for required fact: port_name] ********************** skipping: [junos] ^M TASK [openstack-ml2 : include and run device provider specific tasks] ********** included: /usr/share/ansible/roles/openstack-ml2/tasks/run.yaml for junos ^M TASK [openstack-ml2 : include device role specific tasks] ********************** included: /usr/share/ansible/roles/openstack-ml2/providers/junos/delete_port.yaml for junos ^M TASK [openstack-ml2 : junos: run delete_port using netconf] ******************** included: /usr/share/ansible/roles/openstack-ml2/providers/junos/netconf.yaml for junos ^M TASK [openstack-ml2 : junos: enable netconf if not already enabled] ************ ok: [junos] ^M TASK [openstack-ml2 : junos: set netconf template] ***************************** ok: [junos] ^M TASK [openstack-ml2 : junos: send configuration to device] ********************* An exception occurred during task execution. To see the full traceback, use -vvv. The error was: exclusive^M fatal: [junos]: FAILED! => {"changed": false, "msg": "error editing configuration: \nconfiguration database locked by:\n root terminal (pid 3853) on since 2018-11-22 00:10:35 UTC\n exclusive \n"} ^M PLAY RECAP *********************************************************************^M junos : ok=5 changed=0 unreachable=0 failed=1 ^M : AnsibleRunnerException: ^M Version-Release number of selected component (if applicable): python2-networking-ansible-1.0.4-4 How reproducible: Steps to Reproduce: 1. Create or delete two or more baremetal nodes at once Actual results: One port creation/deletion fails Expected results: Both actions succeed Additional info: The worse scenario is when deleting a node because then vlan configuration is left on the switch port. Next attempt to change the vlan configuration hits the bug 1641963 which leaves the node completely unusable.
Since the problem described in this bug report should be resolved in a recent advisory, it has been closed with a resolution of ERRATA. For information on the advisory, and where to find the updated files, follow the link below. If the solution does not work for you, open a new bug report. https://access.redhat.com/errata/RHEA-2019:0045