)]}'
{"/COMMIT_MSG":[{"author":{"_account_id":14070,"name":"Eric Fried","email":"openstack@fried.cc","username":"efried"},"change_message_id":"0a1aac5a10f8d7f50a7a3e1c92905c6a5153e5ea","unresolved":false,"context_lines":[{"line_number":6,"context_line":""},{"line_number":7,"context_line":"rt: remove misleading log in _remove_deleted_instances_allocations"},{"line_number":8,"context_line":""},{"line_number":9,"context_line":"The ResourceTracker has not \"healed\" allocations since ironic"},{"line_number":10,"context_line":"driver specific support was removed [1] and before that Ocata"},{"line_number":11,"context_line":"compute compat was removed [2]. This change simply removes the"},{"line_number":12,"context_line":"part of the log message saying allocation healing will be skipped"}],"source_content_type":"text/x-gerrit-commit-message","patch_set":1,"id":"7faddb67_e23c7be2","line":9,"range":{"start_line":9,"start_character":20,"end_line":9,"end_character":48},"updated":"2019-07-30 22:39:20.000000000","message":"There\u0027s another message in that block that still talks about healing.\n\nPerhaps we just want to s/healing/deletion of stale allocations/ or something?","commit_id":"b4c478c77c9712c8fddd2d476363900d5668fadf"},{"author":{"_account_id":6873,"name":"Matt Riedemann","email":"mriedem.os@gmail.com","username":"mriedem"},"change_message_id":"7d37616d26f218d556511e4ca4d6c72c733834a9","unresolved":false,"context_lines":[{"line_number":6,"context_line":""},{"line_number":7,"context_line":"rt: remove misleading log in _remove_deleted_instances_allocations"},{"line_number":8,"context_line":""},{"line_number":9,"context_line":"The ResourceTracker has not \"healed\" allocations since ironic"},{"line_number":10,"context_line":"driver specific support was removed [1] and before that Ocata"},{"line_number":11,"context_line":"compute compat was removed [2]. This change simply removes the"},{"line_number":12,"context_line":"part of the log message saying allocation healing will be skipped"}],"source_content_type":"text/x-gerrit-commit-message","patch_set":1,"id":"7faddb67_fae658b0","line":9,"range":{"start_line":9,"start_character":20,"end_line":9,"end_character":48},"in_reply_to":"7faddb67_e23c7be2","updated":"2019-07-31 16:10:26.000000000","message":"I\u0027m thinking I might just abandon this since what this is considering \"healing\" is likely different from what I\u0027m thinking about below, as you\u0027ve pointed out it\u0027s more about deleting stale allocations.","commit_id":"b4c478c77c9712c8fddd2d476363900d5668fadf"}],"nova/compute/resource_tracker.py":[{"author":{"_account_id":14070,"name":"Eric Fried","email":"openstack@fried.cc","username":"efried"},"change_message_id":"0a1aac5a10f8d7f50a7a3e1c92905c6a5153e5ea","unresolved":false,"context_lines":[{"line_number":1379,"context_line":"                # Allocations related to instances being scheduled should not"},{"line_number":1380,"context_line":"                # be deleted if we already wrote the allocation previously."},{"line_number":1381,"context_line":"                LOG.debug(\"Instance %s has been scheduled to this compute \""},{"line_number":1382,"context_line":"                          \"host, the scheduler has made an allocation (%s) \""},{"line_number":1383,"context_line":"                          \"against this compute node but the instance has \""},{"line_number":1384,"context_line":"                          \"yet to start.\", instance_uuid, alloc)"},{"line_number":1385,"context_line":"                continue"}],"source_content_type":"text/x-python","patch_set":1,"id":"7faddb67_02643721","line":1382,"range":{"start_line":1382,"start_character":70,"end_line":1382,"end_character":74},"updated":"2019-07-30 22:39:20.000000000","message":"do we really want a dict in the middle of this message?","commit_id":"b4c478c77c9712c8fddd2d476363900d5668fadf"},{"author":{"_account_id":6873,"name":"Matt Riedemann","email":"mriedem.os@gmail.com","username":"mriedem"},"change_message_id":"7d37616d26f218d556511e4ca4d6c72c733834a9","unresolved":false,"context_lines":[{"line_number":1379,"context_line":"                # Allocations related to instances being scheduled should not"},{"line_number":1380,"context_line":"                # be deleted if we already wrote the allocation previously."},{"line_number":1381,"context_line":"                LOG.debug(\"Instance %s has been scheduled to this compute \""},{"line_number":1382,"context_line":"                          \"host, the scheduler has made an allocation (%s) \""},{"line_number":1383,"context_line":"                          \"against this compute node but the instance has \""},{"line_number":1384,"context_line":"                          \"yet to start.\", instance_uuid, alloc)"},{"line_number":1385,"context_line":"                continue"}],"source_content_type":"text/x-python","patch_set":1,"id":"7faddb67_ba222051","line":1382,"range":{"start_line":1382,"start_character":70,"end_line":1382,"end_character":74},"in_reply_to":"7faddb67_02643721","updated":"2019-07-31 16:10:26.000000000","message":"I didn\u0027t come up with a better way to re-format the message off the top of my head when I pushed this up. Maybe something like:\n\n\"Instance %s has been scheduled to this compute host, the scheduler has made an allocation against this compute node but the instance has yet to start. Allocations: %s\"\n\n?\n\nI also thought about moving instance\u003dinstance for logging instead of putting the instance uuid in the message as a substitution variable but that would be unrelated. Same for logging the compute node / resource provider uuid in the message for context.","commit_id":"b4c478c77c9712c8fddd2d476363900d5668fadf"},{"author":{"_account_id":6873,"name":"Matt Riedemann","email":"mriedem.os@gmail.com","username":"mriedem"},"change_message_id":"7d37616d26f218d556511e4ca4d6c72c733834a9","unresolved":false,"context_lines":[{"line_number":1387,"context_line":"                    instance.node \u003d\u003d cn.hypervisor_hostname):"},{"line_number":1388,"context_line":"                # The instance is supposed to be on this compute host but is"},{"line_number":1389,"context_line":"                # not in the list of actively managed instances."},{"line_number":1390,"context_line":"                LOG.warning(\"Instance %s is not being actively managed by \""},{"line_number":1391,"context_line":"                            \"this compute host but has allocations \""},{"line_number":1392,"context_line":"                            \"referencing this compute host: %s. Skipping \""},{"line_number":1393,"context_line":"                            \"heal of allocation because we do not know \""}],"source_content_type":"text/x-python","patch_set":1,"id":"7faddb67_ba7ea016","line":1390,"range":{"start_line":1390,"start_character":20,"end_line":1390,"end_character":27},"updated":"2019-07-31 16:10:26.000000000","message":"Unrelated, but this shows up a lot in CI runs:\n\nhttp://logstash.openstack.org/#dashboard/file/logstash.json?query\u003dmessage%3A%5C%22is%20not%20being%20actively%20managed%20by%5C%22%20AND%20tags%3A%5C%22screen-n-cpu.txt%5C%22\u0026from\u003d7d\n\nWhich indicates it\u0027s normal and shouldn\u0027t be a warning. Likely racing with some migration operation.\n\nActually digging into one of these shows it\u0027s hitting this while unshelving an instance. During unshelve we call instance_claim which will set the instance.host and instance.node values:\n\nhttps://github.com/openstack/nova/blob/619c0c676aae5359225c54bc27ce349e138e420e/nova/compute/resource_tracker.py#L208\n\nbefore eventually adding the instance to the tracked_instances dict:\n\nhttps://github.com/openstack/nova/blob/619c0c676aae5359225c54bc27ce349e138e420e/nova/compute/resource_tracker.py#L217\n\nIf update_available_resource periodic runs between those times we\u0027ll get this warning. So this should probably not be a warning if the instance task_state is set (because the instance is undergoing some state transition). Arguably we don\u0027t really need to log this (or make it debug) if we know the instance is being unshelved (task_state\u003dspawning and shelved_at is in instance.system_metadata - but maybe that\u0027s going a bit too far).\n\nAnyway, that\u0027s a separate bug.","commit_id":"b4c478c77c9712c8fddd2d476363900d5668fadf"},{"author":{"_account_id":6873,"name":"Matt Riedemann","email":"mriedem.os@gmail.com","username":"mriedem"},"change_message_id":"dc6c9fade40ce7fb7f86435da0f74cbcb6551acf","unresolved":false,"context_lines":[{"line_number":1387,"context_line":"                    instance.node \u003d\u003d cn.hypervisor_hostname):"},{"line_number":1388,"context_line":"                # The instance is supposed to be on this compute host but is"},{"line_number":1389,"context_line":"                # not in the list of actively managed instances."},{"line_number":1390,"context_line":"                LOG.warning(\"Instance %s is not being actively managed by \""},{"line_number":1391,"context_line":"                            \"this compute host but has allocations \""},{"line_number":1392,"context_line":"                            \"referencing this compute host: %s. Skipping \""},{"line_number":1393,"context_line":"                            \"heal of allocation because we do not know \""}],"source_content_type":"text/x-python","patch_set":1,"id":"7faddb67_d53ecdb6","line":1390,"range":{"start_line":1390,"start_character":20,"end_line":1390,"end_character":27},"in_reply_to":"7faddb67_ba7ea016","updated":"2019-07-31 16:21:58.000000000","message":"I\u0027ve reported bug 1838541 for this.","commit_id":"b4c478c77c9712c8fddd2d476363900d5668fadf"},{"author":{"_account_id":14070,"name":"Eric Fried","email":"openstack@fried.cc","username":"efried"},"change_message_id":"0a1aac5a10f8d7f50a7a3e1c92905c6a5153e5ea","unresolved":false,"context_lines":[{"line_number":1389,"context_line":"                # not in the list of actively managed instances."},{"line_number":1390,"context_line":"                LOG.warning(\"Instance %s is not being actively managed by \""},{"line_number":1391,"context_line":"                            \"this compute host but has allocations \""},{"line_number":1392,"context_line":"                            \"referencing this compute host: %s. Skipping \""},{"line_number":1393,"context_line":"                            \"heal of allocation because we do not know \""},{"line_number":1394,"context_line":"                            \"what to do.\", instance_uuid, alloc)"},{"line_number":1395,"context_line":"                continue"},{"line_number":1396,"context_line":"            if instance.host !\u003d cn.host:"}],"source_content_type":"text/x-python","patch_set":1,"id":"7faddb67_822587c4","line":1393,"range":{"start_line":1392,"start_character":64,"end_line":1393,"end_character":47},"updated":"2019-07-30 22:39:20.000000000","message":"Does this reference to \u0027heal\u0027 also no longer make sense?","commit_id":"b4c478c77c9712c8fddd2d476363900d5668fadf"},{"author":{"_account_id":6873,"name":"Matt Riedemann","email":"mriedem.os@gmail.com","username":"mriedem"},"change_message_id":"7d37616d26f218d556511e4ca4d6c72c733834a9","unresolved":false,"context_lines":[{"line_number":1389,"context_line":"                # not in the list of actively managed instances."},{"line_number":1390,"context_line":"                LOG.warning(\"Instance %s is not being actively managed by \""},{"line_number":1391,"context_line":"                            \"this compute host but has allocations \""},{"line_number":1392,"context_line":"                            \"referencing this compute host: %s. Skipping \""},{"line_number":1393,"context_line":"                            \"heal of allocation because we do not know \""},{"line_number":1394,"context_line":"                            \"what to do.\", instance_uuid, alloc)"},{"line_number":1395,"context_line":"                continue"},{"line_number":1396,"context_line":"            if instance.host !\u003d cn.host:"}],"source_content_type":"text/x-python","patch_set":1,"id":"7faddb67_ba954042","line":1393,"range":{"start_line":1392,"start_character":64,"end_line":1393,"end_character":47},"in_reply_to":"7faddb67_822587c4","updated":"2019-07-31 16:10:26.000000000","message":"Maybe...if it does make sense then that means my change doesn\u0027t really make sense, and if we consider \u0027healing allocations\u0027 just not deleting them for deleted instances - rather than what I\u0027m thinking \u0027healing\u0027 means, which was the old stuff the RT used to do as noted in the commit message, then I could just abandon this change.","commit_id":"b4c478c77c9712c8fddd2d476363900d5668fadf"}]}