)]}'
{"roles/tripleo-inventory/tasks/inventory.yml":[{"author":{"_account_id":23811,"name":"Oliver Walsh","email":"owalsh@redhat.com","username":"owalsh"},"change_message_id":"5a93b8aa0334529724b85c1e670300d45b09bbe7","unresolved":false,"context_lines":[{"line_number":59,"context_line":"- name: set_fact for undercloud ip"},{"line_number":60,"context_line":"  set_fact:"},{"line_number":61,"context_line":"    undercloud_ip: \"{{ hostvars[\u0027undercloud\u0027].undercloud_ip }}\""},{"line_number":62,"context_line":"    cacheable: false"},{"line_number":63,"context_line":"  when: hostvars[\u0027undercloud\u0027] is defined and hostvars[\u0027undercloud\u0027].undercloud_ip is defined"},{"line_number":64,"context_line":""},{"line_number":65,"context_line":"- name: set_fact for virthost"}],"source_content_type":"text/x-yaml","patch_set":1,"id":"7faddb67_ddb6b689","line":62,"range":{"start_line":62,"start_character":4,"end_line":62,"end_character":20},"updated":"2019-09-05 08:53:14.000000000","message":"This works for local dev... also had to stop caching dlrn hash etc... as the current-tripleo--rdo link changed between runs...\n\nHowever I was scratching my head wondering why the hell this was done in the first place. Surely it can\u0027t be this straightforward...\nThe commit msg in the patch (https://review.opendev.org/525237) gives no hints whatsoever, but in the review comments trown says:\n\n    I think this actually solves the use case we need it for. \n    We only need the set_facts from one playbook run to \n    persist at the same precedence in another playbook run. \n    This change accomplishes that.\n\n    We are currently working on a sprint to allow running \n    individual playbooks from the CI code so as to use it to \n    setup development environments. If we overcomplicate this \n    by creating custom facts modules that will almost \n    certainly mean we do not accomplish that goal this sprint.\n\n    If there is some serious downside to what we are doing, \n    obviously that is worth delaying for. However, I am not \n    seeing that.\n\nSo yea, there is a downside!","commit_id":"75ed9a740c20f8f54e949fb9c670582587ad98ef"},{"author":{"_account_id":23811,"name":"Oliver Walsh","email":"owalsh@redhat.com","username":"owalsh"},"change_message_id":"fac668a84cc3c908c6dd7e989217ad3a13cf4c45","unresolved":false,"context_lines":[{"line_number":59,"context_line":"- name: set_fact for undercloud ip"},{"line_number":60,"context_line":"  set_fact:"},{"line_number":61,"context_line":"    undercloud_ip: \"{{ hostvars[\u0027undercloud\u0027].undercloud_ip }}\""},{"line_number":62,"context_line":"    cacheable: false"},{"line_number":63,"context_line":"  when: hostvars[\u0027undercloud\u0027] is defined and hostvars[\u0027undercloud\u0027].undercloud_ip is defined"},{"line_number":64,"context_line":""},{"line_number":65,"context_line":"- name: set_fact for virthost"}],"source_content_type":"text/x-yaml","patch_set":1,"id":"7faddb67_981ccc13","line":62,"range":{"start_line":62,"start_character":4,"end_line":62,"end_character":20},"in_reply_to":"7faddb67_ddb6b689","updated":"2019-09-05 09:12:19.000000000","message":"Why not just add a flag to control the fact caching? e.g --retain-facts similar to --retain-inventory","commit_id":"75ed9a740c20f8f54e949fb9c670582587ad98ef"}]}