)]}'
{"/PATCHSET_LEVEL":[{"author":{"_account_id":28223,"name":"Cedric Jeanneret","display_name":"cjeanner (Tengu)","email":"cjeanner@redhat.com","username":"cjeanner"},"change_message_id":"4ed4868817fe83f5e796dc5c007d9283fa1ca85d","unresolved":false,"context_lines":[],"source_content_type":"","patch_set":1,"id":"065deaa0_61452447","updated":"2022-02-07 06:16:12.000000000","message":"FYI, there are related tasks in another file:\ncommon/common-container-setup-tasks.yaml\nhttps://opendev.org/openstack/tripleo-heat-templates/src/branch/master/common/common-container-setup-tasks.yaml#L20-L53\n\nIMHO we should put that cleanup task in that same common-container-setup-tasks since we will eventually face the same issue on the overcloud nodes. WDYT?\n\nnote: -1 in order to raise awareness and engage discussion - the idea itself is more than good ;).","commit_id":"eb738fa8bab0cca63c837942a48357de85307bd0"},{"author":{"_account_id":26343,"name":"Jose Luis Franco","email":"jfrancoa@redhat.com","username":"jfrancoa"},"change_message_id":"cc8fe87b463781d0d772ecd90eebbfd387acb9a2","unresolved":false,"context_lines":[],"source_content_type":"","patch_set":1,"id":"a26fcb12_2373ccff","updated":"2022-02-07 13:21:32.000000000","message":"I\u0027ll dig into what @Rabi mentioned in his latest comment. It looks like a better solution, to fix the bug than remove the whole folder, to be honest. I\u0027ll abandon this patch once I get a fix in place.\n\nThanks for the feedback","commit_id":"eb738fa8bab0cca63c837942a48357de85307bd0"},{"author":{"_account_id":8833,"name":"Rabi Mishra","email":"ramishra@redhat.com","username":"rabi"},"change_message_id":"08cdcb1b0c83f92ee7c0e9bacfddb3e54f0f729f","unresolved":false,"context_lines":[],"source_content_type":"","patch_set":1,"id":"f7ffb366_d3c60b1e","updated":"2022-02-07 07:16:54.000000000","message":"What is the current issue? I don\u0027t see enough details in the LP bug. Why it happens for only nova/glance and not for mistral/zaqar etc (Not seen in any upstream upgrade jobs). I thought we cleanup/recreate the configs every time[1].\n\n[1] https://github.com/openstack/tripleo-ansible/blob/master/tripleo_ansible/ansible_plugins/modules/container_startup_config.py#L93-L102","commit_id":"eb738fa8bab0cca63c837942a48357de85307bd0"},{"author":{"_account_id":6926,"name":"Bogdan Dobrelya","email":"bdobreli@redhat.com","username":"bogdando"},"change_message_id":"3fec5676530eea21295cae35842f46b606a3b648","unresolved":false,"context_lines":[],"source_content_type":"","patch_set":1,"id":"8bf6ffad_55e891c5","updated":"2022-02-08 14:32:28.000000000","message":"it seems we should stick with a Wallaby backport of https://review.opendev.org/c/openstack/tripleo-ansible/+/828212","commit_id":"eb738fa8bab0cca63c837942a48357de85307bd0"},{"author":{"_account_id":8833,"name":"Rabi Mishra","email":"ramishra@redhat.com","username":"rabi"},"change_message_id":"3f64aa5cd204238bfa41d9e492968edc696b2516","unresolved":false,"context_lines":[],"source_content_type":"","patch_set":1,"id":"92935d56_72638f12","in_reply_to":"51f82cea_aafeaf9b","updated":"2022-02-07 12:47:52.000000000","message":"OK, as step5 is missing from the config-data[1] that directory is not cleaned up/recreated. I think we should fix it in that ansible module[2] (may be check for additional step folders present).\n\n\n[1] https://github.com/openstack/tripleo-heat-templates/blob/stable/wallaby/common/common-container-setup-tasks.yaml#L76\n\n[2] https://github.com/openstack/tripleo-ansible/blob/master/tripleo_ansible/ansible_plugins/modules/container_startup_config.py#L93-L102","commit_id":"eb738fa8bab0cca63c837942a48357de85307bd0"},{"author":{"_account_id":8833,"name":"Rabi Mishra","email":"ramishra@redhat.com","username":"rabi"},"change_message_id":"e18875341cb00a3d7c01389b09367e33e105282a","unresolved":false,"context_lines":[],"source_content_type":"","patch_set":1,"id":"f2f7bb39_77b244a5","in_reply_to":"79af237c_20e22410","updated":"2022-02-08 03:38:18.000000000","message":"Let me know if https://review.opendev.org/c/openstack/tripleo-ansible/+/828212 helps?","commit_id":"eb738fa8bab0cca63c837942a48357de85307bd0"},{"author":{"_account_id":6926,"name":"Bogdan Dobrelya","email":"bdobreli@redhat.com","username":"bogdando"},"change_message_id":"1fca92821454d3853bf78942e1a02a4ee7c1a423","unresolved":false,"context_lines":[],"source_content_type":"","patch_set":1,"id":"f5a40820_167fdcec","in_reply_to":"92935d56_72638f12","updated":"2022-02-07 14:30:07.000000000","message":"\u003e OK, as step5 is missing from the config-data[1] that directory is not cleaned up/recreated. I think we should fix it in that ansible module[2] (may be check for additional step folders present).\n\u003e \n\u003e \n\u003e [1] https://github.com/openstack/tripleo-heat-templates/blob/stable/wallaby/common/common-container-setup-tasks.yaml#L76\n\u003e \n\u003e [2] https://github.com/openstack/tripleo-ansible/blob/master/tripleo_ansible/ansible_plugins/modules/container_startup_config.py#L93-L102\n\nI vote to just remove it in tht, please don\u0027t make container manager aware of upgrade dance and legacy formats. That would only make it slowly to execure","commit_id":"eb738fa8bab0cca63c837942a48357de85307bd0"},{"author":{"_account_id":8833,"name":"Rabi Mishra","email":"ramishra@redhat.com","username":"rabi"},"change_message_id":"44f874fd7c2a8a3f0a4c3ef8f0eca98c4e0f905c","unresolved":false,"context_lines":[],"source_content_type":"","patch_set":1,"id":"79af237c_20e22410","in_reply_to":"f5a40820_167fdcec","updated":"2022-02-07 14:40:16.000000000","message":"The module has to intelligent enough to handle existing configs i.e weather to overwrite them or not etc.. \n\nHow we handle the configs for steps missing during upgrade can be in THT if we want (not blindly cleanup everything). It\u0027s just putting code in multiple places makes it difficult to follow.","commit_id":"eb738fa8bab0cca63c837942a48357de85307bd0"},{"author":{"_account_id":26343,"name":"Jose Luis Franco","email":"jfrancoa@redhat.com","username":"jfrancoa"},"change_message_id":"642e11d21e456ff9cc9ce0836dec44b27c7c03a3","unresolved":false,"context_lines":[],"source_content_type":"","patch_set":1,"id":"51f82cea_aafeaf9b","in_reply_to":"f7ffb366_d3c60b1e","updated":"2022-02-07 11:04:19.000000000","message":"Hi @Rabi, maybe you can get more insights from https://bugzilla.redhat.com/show_bug.cgi?id\u003d2051376 . The thing is that maybe this fix is like killing a fly with a cannon, but basically the container tripleo configs for step 5 remain after the UC upgrade (while for any other step the configs belonging to deprecated services get removed). This causes the deployment tasks to start up nova_compute, nova_wait_for_compute_service and mistral_db_populate containers during the OSP17 deployment steps. And as nova_wait_for_compute service expects nova_conductor to be up, the upgrade fails...but none of these three services should be starting up.\nThis fix solves the issue and the container\u0027s configs get recreated, but maybe we should try to address the problem per se and find out why only those configs remain. WDYT?","commit_id":"eb738fa8bab0cca63c837942a48357de85307bd0"}]}