)]}'
{"/COMMIT_MSG":[{"author":{"_account_id":9708,"name":"Balazs Gibizer","display_name":"gibi","email":"gibizer@gmail.com","username":"gibi"},"change_message_id":"dc6283bf87da15196e9f589421042c40d0c6bb1f","unresolved":true,"context_lines":[{"line_number":12,"context_line":"config, further migrations will update the domain name to the new"},{"line_number":13,"context_line":"format."},{"line_number":14,"context_line":""},{"line_number":15,"context_line":"This is a seamless operation that happens only on the source node"},{"line_number":16,"context_line":"and does not require communication between the two and thus does"},{"line_number":17,"context_line":"not require an RPC or service version bump. Supported by libvirt"},{"line_number":18,"context_line":"since ~2013."}],"source_content_type":"text/x-gerrit-commit-message","patch_set":1,"id":"8fa5d03b_8d5d4870","line":15,"range":{"start_line":15,"start_character":0,"end_line":15,"end_character":65},"updated":"2026-10-26 10:48:00.000000000","message":"yeah the calculation of the new name happens on the source node based on the source node\u0027s config, but then libvirt applies the new name on the dest node only. This at least requires coordinated changes of the config option across computes. Otherwise we will have a strange thing e.g. computeA having a UUID based template configured so it boots VMs with UUID names, but also all VMs moved *away* from computeA will have a UUID name on whatever dest node they land on. So if the operator does not want a mess they should change the config to UUID on all compute within the same cell / AZ or whatever set of compute they tend to live migrate across.","commit_id":"696c2882e5a25fee959890365ead9e08258a81d2"},{"author":{"_account_id":9708,"name":"Balazs Gibizer","display_name":"gibi","email":"gibizer@gmail.com","username":"gibi"},"change_message_id":"13009eae697ea9e84e81a20b4cd527fb5a71108c","unresolved":false,"context_lines":[{"line_number":12,"context_line":"config, further migrations will update the domain name to the new"},{"line_number":13,"context_line":"format."},{"line_number":14,"context_line":""},{"line_number":15,"context_line":"This is a seamless operation that happens only on the source node"},{"line_number":16,"context_line":"and does not require communication between the two and thus does"},{"line_number":17,"context_line":"not require an RPC or service version bump. Supported by libvirt"},{"line_number":18,"context_line":"since ~2013."}],"source_content_type":"text/x-gerrit-commit-message","patch_set":1,"id":"367bc2bb_639656d3","line":15,"range":{"start_line":15,"start_character":0,"end_line":15,"end_character":65},"in_reply_to":"4b688724_9133ce19","updated":"2026-10-27 17:18:46.000000000","message":"Cool having an instance record already is good. Then the next boot succeeds.","commit_id":"696c2882e5a25fee959890365ead9e08258a81d2"},{"author":{"_account_id":9708,"name":"Balazs Gibizer","display_name":"gibi","email":"gibizer@gmail.com","username":"gibi"},"change_message_id":"808af30e9be4b54ec9743cbf1735296a04c903ce","unresolved":true,"context_lines":[{"line_number":12,"context_line":"config, further migrations will update the domain name to the new"},{"line_number":13,"context_line":"format."},{"line_number":14,"context_line":""},{"line_number":15,"context_line":"This is a seamless operation that happens only on the source node"},{"line_number":16,"context_line":"and does not require communication between the two and thus does"},{"line_number":17,"context_line":"not require an RPC or service version bump. Supported by libvirt"},{"line_number":18,"context_line":"since ~2013."}],"source_content_type":"text/x-gerrit-commit-message","patch_set":1,"id":"bb34f6d1_63fd82da","line":15,"range":{"start_line":15,"start_character":0,"end_line":15,"end_character":65},"in_reply_to":"8fa5d03b_8d5d4870","updated":"2026-10-26 10:53:52.000000000","message":"Let\u0027s address this in the next commit.","commit_id":"696c2882e5a25fee959890365ead9e08258a81d2"},{"author":{"_account_id":4393,"name":"Dan Smith","email":"dms@danplanet.com","username":"danms"},"change_message_id":"d46e74f92db02a26da2f813941917847bdbab819","unresolved":true,"context_lines":[{"line_number":12,"context_line":"config, further migrations will update the domain name to the new"},{"line_number":13,"context_line":"format."},{"line_number":14,"context_line":""},{"line_number":15,"context_line":"This is a seamless operation that happens only on the source node"},{"line_number":16,"context_line":"and does not require communication between the two and thus does"},{"line_number":17,"context_line":"not require an RPC or service version bump. Supported by libvirt"},{"line_number":18,"context_line":"since ~2013."}],"source_content_type":"text/x-gerrit-commit-message","patch_set":1,"id":"fbf0e4ec_55debcf4","line":15,"range":{"start_line":15,"start_character":0,"end_line":15,"end_character":65},"in_reply_to":"bb34f6d1_63fd82da","updated":"2026-10-26 13:54:01.000000000","message":"Today if you have your computes configured differently then you will end up with instances on a compute with a name that differs from what is configured (if it was migrated in from a compute that has a different template). The only actual difference here is that an instance can go from the original template to the new one during a migration. But from a higher level \"the instance name is whatever the source had configured\" remains the same. There\u0027s just a delta if you\u0027ve changed the source config after instance boot and before migration.\n\nThe good news is that the work we did to stop using names for anything means that all of this is harmless. I guess I assume it\u0027s implied good practice to use the same template everywhere, but it\u0027s by no means required anymore. Theoretically someone could have `instance-%(host)s-%(id)s` and that would be unique enough for the cells thing, and then they could come to their senses later and take the uuid default. All of that can coexist an they could not change their source nodes until they have emptied a cell if they want.","commit_id":"696c2882e5a25fee959890365ead9e08258a81d2"},{"author":{"_account_id":4393,"name":"Dan Smith","email":"dms@danplanet.com","username":"danms"},"change_message_id":"116b568a336e02e3fadc35d8f72e6f94b7f998c8","unresolved":true,"context_lines":[{"line_number":12,"context_line":"config, further migrations will update the domain name to the new"},{"line_number":13,"context_line":"format."},{"line_number":14,"context_line":""},{"line_number":15,"context_line":"This is a seamless operation that happens only on the source node"},{"line_number":16,"context_line":"and does not require communication between the two and thus does"},{"line_number":17,"context_line":"not require an RPC or service version bump. Supported by libvirt"},{"line_number":18,"context_line":"since ~2013."}],"source_content_type":"text/x-gerrit-commit-message","patch_set":1,"id":"4b688724_9133ce19","line":15,"range":{"start_line":15,"start_character":0,"end_line":15,"end_character":65},"in_reply_to":"f2f015bd_d3b17271","updated":"2026-10-27 15:39:46.000000000","message":"That\u0027s a fair concern and one I hadn\u0027t really thought of (i.e. migrating in instance-002, where 2 is the next instance id that would be created). However, since we create the instance record before we ever try to boot, that would only (AFAICT) create a blocker for a single instance and not a \"deadlock\" type situation.\n\nThat said, it\u0027s (another) great reason to NOT base the instance name on the internal DB ID. Hopefully the extra text I added to the next commit is helpful, but perhaps it should be strengthened further to address your concern.","commit_id":"696c2882e5a25fee959890365ead9e08258a81d2"},{"author":{"_account_id":9708,"name":"Balazs Gibizer","display_name":"gibi","email":"gibizer@gmail.com","username":"gibi"},"change_message_id":"60386f2a49a0da8a7cf2ec8a1627251ef480d568","unresolved":true,"context_lines":[{"line_number":12,"context_line":"config, further migrations will update the domain name to the new"},{"line_number":13,"context_line":"format."},{"line_number":14,"context_line":""},{"line_number":15,"context_line":"This is a seamless operation that happens only on the source node"},{"line_number":16,"context_line":"and does not require communication between the two and thus does"},{"line_number":17,"context_line":"not require an RPC or service version bump. Supported by libvirt"},{"line_number":18,"context_line":"since ~2013."}],"source_content_type":"text/x-gerrit-commit-message","patch_set":1,"id":"f2f015bd_d3b17271","line":15,"range":{"start_line":15,"start_character":0,"end_line":15,"end_character":65},"in_reply_to":"fbf0e4ec_55debcf4","updated":"2026-10-27 15:29:29.000000000","message":"OK. If this does not lead to any deadlock situation like migrating an instance to a  host with a name that e.g. block the next VM to be booted on that host then I\u0027m fine. I\u0027m OK if the migration operation fails (cleanly) due to name conflict. But I want to avoid that a migration succeeds but then nothing can be booted to the dest host any more.","commit_id":"696c2882e5a25fee959890365ead9e08258a81d2"},{"author":{"_account_id":9708,"name":"Balazs Gibizer","display_name":"gibi","email":"gibizer@gmail.com","username":"gibi"},"change_message_id":"dc6283bf87da15196e9f589421042c40d0c6bb1f","unresolved":true,"context_lines":[{"line_number":16,"context_line":"and does not require communication between the two and thus does"},{"line_number":17,"context_line":"not require an RPC or service version bump. Supported by libvirt"},{"line_number":18,"context_line":"since ~2013."},{"line_number":19,"context_line":""},{"line_number":20,"context_line":"Change-Id: Idff9da65b39a94b6fee88e5d4da9137a2fb07faa"},{"line_number":21,"context_line":"Signed-off-by: Dan Smith \u003cdansmith@redhat.com\u003e"}],"source_content_type":"text/x-gerrit-commit-message","patch_set":1,"id":"eea25bba_757739ed","line":19,"updated":"2026-10-26 10:48:00.000000000","message":"We need to somehow tell our users when it is safe to do this now as the config doc still says it is unsafe if you have instances https://docs.openstack.org/nova/latest/configuration/config.html#DEFAULT.instance_name_template","commit_id":"696c2882e5a25fee959890365ead9e08258a81d2"},{"author":{"_account_id":9708,"name":"Balazs Gibizer","display_name":"gibi","email":"gibizer@gmail.com","username":"gibi"},"change_message_id":"808af30e9be4b54ec9743cbf1735296a04c903ce","unresolved":false,"context_lines":[{"line_number":16,"context_line":"and does not require communication between the two and thus does"},{"line_number":17,"context_line":"not require an RPC or service version bump. Supported by libvirt"},{"line_number":18,"context_line":"since ~2013."},{"line_number":19,"context_line":""},{"line_number":20,"context_line":"Change-Id: Idff9da65b39a94b6fee88e5d4da9137a2fb07faa"},{"line_number":21,"context_line":"Signed-off-by: Dan Smith \u003cdansmith@redhat.com\u003e"}],"source_content_type":"text/x-gerrit-commit-message","patch_set":1,"id":"a2afb01f_7e93cd0e","line":19,"in_reply_to":"eea25bba_757739ed","updated":"2026-10-26 10:53:52.000000000","message":"fixed in the next commit.","commit_id":"696c2882e5a25fee959890365ead9e08258a81d2"}],"/PATCHSET_LEVEL":[{"author":{"_account_id":9708,"name":"Balazs Gibizer","display_name":"gibi","email":"gibizer@gmail.com","username":"gibi"},"change_message_id":"13009eae697ea9e84e81a20b4cd527fb5a71108c","unresolved":false,"context_lines":[],"source_content_type":"","patch_set":1,"id":"7c8dfed4_8a6b15a0","updated":"2026-10-27 17:18:46.000000000","message":"I did some local testing of the upgrade of this feature with pre-existing VMs and live migrations back and forth after the upgrade. All looks solid.","commit_id":"696c2882e5a25fee959890365ead9e08258a81d2"},{"author":{"_account_id":9708,"name":"Balazs Gibizer","display_name":"gibi","email":"gibizer@gmail.com","username":"gibi"},"change_message_id":"dc6283bf87da15196e9f589421042c40d0c6bb1f","unresolved":false,"context_lines":[],"source_content_type":"","patch_set":1,"id":"2166ab24_7e4794a9","updated":"2026-10-26 10:48:00.000000000","message":"I feel like we need a bit of doc around this behavior as it can be complicated to predict what happens if the template config is changed on a compute node.","commit_id":"696c2882e5a25fee959890365ead9e08258a81d2"},{"author":{"_account_id":4393,"name":"Dan Smith","email":"dms@danplanet.com","username":"danms"},"change_message_id":"d46e74f92db02a26da2f813941917847bdbab819","unresolved":false,"context_lines":[],"source_content_type":"","patch_set":1,"id":"a375e635_81319c59","updated":"2026-10-26 13:54:01.000000000","message":"Thanks for close review gibi!","commit_id":"696c2882e5a25fee959890365ead9e08258a81d2"},{"author":{"_account_id":4393,"name":"Dan Smith","email":"dms@danplanet.com","username":"danms"},"change_message_id":"a6b98f6309de632e9f04956ceb640d97ab30e5a0","unresolved":false,"context_lines":[],"source_content_type":"","patch_set":1,"id":"832a946f_1ec05993","updated":"2026-10-29 14:51:36.000000000","message":"recheck build timeout","commit_id":"696c2882e5a25fee959890365ead9e08258a81d2"},{"author":{"_account_id":4393,"name":"Dan Smith","email":"dms@danplanet.com","username":"danms"},"change_message_id":"aa9c40cbce7bd59e5825d2120b35938ba17cc700","unresolved":false,"context_lines":[],"source_content_type":"","patch_set":1,"id":"57505549_bdbd9d14","updated":"2026-10-20 14:11:05.000000000","message":"recheck something broke midway through and we stopped getting neutron events for any new instances thus plugging always failed","commit_id":"696c2882e5a25fee959890365ead9e08258a81d2"},{"author":{"_account_id":9708,"name":"Balazs Gibizer","display_name":"gibi","email":"gibizer@gmail.com","username":"gibi"},"change_message_id":"12407b5c65655002b84afc97a3f657307489c1cf","unresolved":false,"context_lines":[],"source_content_type":"","patch_set":1,"id":"19de3ffd_3c405d21","updated":"2026-10-26 10:25:34.000000000","message":"recheck the tempest failures are mostly due to vif plug timeout\n``\nOct 20 15:20:11.947460 npf12ce572dbf34 nova-compute[74847]: WARNING nova.compute.manager [None req-2723e489-d4b7-428d-a75f-022e2de1fcf9 tempest-AttachVolumeNegativeTest-1559270162 tempest-AttachVolumeNegativeTest-1559270162-project-member] [instance: 94424ce2-5815-4eeb-97a9-44e380faa736] Timeout waiting for [\u0027network-vif-plugged-c5e1fae6-16e8-45ef-b83c-e38e9e791d27\u0027] for instance with vm_state building and task_state spawning. Event states are: network-vif-plugged-c5e1fae6-16e8-45ef-b83c-e38e9e791d27: timed out after 300.00 seconds: nova.exception.InstanceEventTimeout\n```","commit_id":"696c2882e5a25fee959890365ead9e08258a81d2"},{"author":{"_account_id":9708,"name":"Balazs Gibizer","display_name":"gibi","email":"gibizer@gmail.com","username":"gibi"},"change_message_id":"808af30e9be4b54ec9743cbf1735296a04c903ce","unresolved":false,"context_lines":[],"source_content_type":"","patch_set":1,"id":"787bce4e_f54412b4","updated":"2026-10-26 10:53:52.000000000","message":"saw the other commit addressing my comments. So I switched my vote to positive. I will run some local testing with the two patches before upgrading to +2.","commit_id":"696c2882e5a25fee959890365ead9e08258a81d2"}],"nova/conf/compute.py":[{"author":{"_account_id":9708,"name":"Balazs Gibizer","display_name":"gibi","email":"gibizer@gmail.com","username":"gibi"},"change_message_id":"dc6283bf87da15196e9f589421042c40d0c6bb1f","unresolved":true,"context_lines":[{"line_number":1564,"context_line":"is *not* the display name you enter when creating an instance (via Horizon"},{"line_number":1565,"context_line":"or CLI). For a new deployment it is advisable to change the default value"},{"line_number":1566,"context_line":"(which uses the database autoincrement) to another value which makes use"},{"line_number":1567,"context_line":"of the attributes of an instance, like ``instance-%(uuid)s``. If you"},{"line_number":1568,"context_line":"already have instances in your deployment when you change this, your"},{"line_number":1569,"context_line":"deployment will break."},{"line_number":1570,"context_line":""},{"line_number":1571,"context_line":"Possible values:"},{"line_number":1572,"context_line":""}],"source_content_type":"text/x-python","patch_set":1,"id":"2420e33e_4d766f59","line":1569,"range":{"start_line":1567,"start_character":62,"end_line":1569,"end_character":22},"updated":"2026-10-26 10:48:00.000000000","message":"this is a bit contradicting now with the commit message that suggest that changing the config actually OK.","commit_id":"696c2882e5a25fee959890365ead9e08258a81d2"},{"author":{"_account_id":9708,"name":"Balazs Gibizer","display_name":"gibi","email":"gibizer@gmail.com","username":"gibi"},"change_message_id":"808af30e9be4b54ec9743cbf1735296a04c903ce","unresolved":false,"context_lines":[{"line_number":1564,"context_line":"is *not* the display name you enter when creating an instance (via Horizon"},{"line_number":1565,"context_line":"or CLI). For a new deployment it is advisable to change the default value"},{"line_number":1566,"context_line":"(which uses the database autoincrement) to another value which makes use"},{"line_number":1567,"context_line":"of the attributes of an instance, like ``instance-%(uuid)s``. If you"},{"line_number":1568,"context_line":"already have instances in your deployment when you change this, your"},{"line_number":1569,"context_line":"deployment will break."},{"line_number":1570,"context_line":""},{"line_number":1571,"context_line":"Possible values:"},{"line_number":1572,"context_line":""}],"source_content_type":"text/x-python","patch_set":1,"id":"8a5db731_a01243a0","line":1569,"range":{"start_line":1567,"start_character":62,"end_line":1569,"end_character":22},"in_reply_to":"2420e33e_4d766f59","updated":"2026-10-26 10:53:52.000000000","message":"I see it being fixed in the next commit.","commit_id":"696c2882e5a25fee959890365ead9e08258a81d2"},{"author":{"_account_id":9708,"name":"Balazs Gibizer","display_name":"gibi","email":"gibizer@gmail.com","username":"gibi"},"change_message_id":"dc6283bf87da15196e9f589421042c40d0c6bb1f","unresolved":false,"context_lines":[{"line_number":1572,"context_line":""},{"line_number":1573,"context_line":"* A string which either uses the instance database ID (like the"},{"line_number":1574,"context_line":"  default)"},{"line_number":1575,"context_line":"* A string with a list of named database columns, for example ``%(id)d``"},{"line_number":1576,"context_line":"  or ``%(uuid)s`` or ``%(hostname)s``."},{"line_number":1577,"context_line":"\"\"\"),"},{"line_number":1578,"context_line":"]"}],"source_content_type":"text/x-python","patch_set":1,"id":"b25ed1e4_ad10c418","line":1575,"range":{"start_line":1575,"start_character":18,"end_line":1575,"end_character":48},"updated":"2026-10-26 10:48:00.000000000","message":"that is also incorrect, the implementation uses object field names not DB column names. But this is probably a different commit to fix.","commit_id":"696c2882e5a25fee959890365ead9e08258a81d2"},{"author":{"_account_id":4393,"name":"Dan Smith","email":"dms@danplanet.com","username":"danms"},"change_message_id":"d46e74f92db02a26da2f813941917847bdbab819","unresolved":false,"context_lines":[{"line_number":1572,"context_line":""},{"line_number":1573,"context_line":"* A string which either uses the instance database ID (like the"},{"line_number":1574,"context_line":"  default)"},{"line_number":1575,"context_line":"* A string with a list of named database columns, for example ``%(id)d``"},{"line_number":1576,"context_line":"  or ``%(uuid)s`` or ``%(hostname)s``."},{"line_number":1577,"context_line":"\"\"\"),"},{"line_number":1578,"context_line":"]"}],"source_content_type":"text/x-python","patch_set":1,"id":"c94b9b2b_b6966159","line":1575,"range":{"start_line":1575,"start_character":18,"end_line":1575,"end_character":48},"in_reply_to":"b25ed1e4_ad10c418","updated":"2026-10-26 13:54:01.000000000","message":"Yeah, this has been wrong for quite a while now","commit_id":"696c2882e5a25fee959890365ead9e08258a81d2"}]}