)]}'
{"/COMMIT_MSG":[{"author":{"_account_id":11604,"name":"sean mooney","email":"smooney@redhat.com","username":"sean-k-mooney"},"change_message_id":"1d07408c047d3eee64ab6a9f25648c0d907a7f9a","unresolved":true,"context_lines":[{"line_number":11,"context_line":"Re-instantiate the key manager on every call."},{"line_number":12,"context_line":""},{"line_number":13,"context_line":"Related-Bug: #1873742"},{"line_number":14,"context_line":"Related-Bug: #2086765"},{"line_number":15,"context_line":"Change-Id: I5ef22201df3aab259c49c6ad908a63e645d83556"}],"source_content_type":"text/x-gerrit-commit-message","patch_set":3,"id":"eb2a0a1c_a7ef67df","line":14,"updated":"2024-11-11 03:10:16.000000000","message":"there is no explanation as to why either bug would apply to nova.\n\nunless you can explain why this is required and ideally create a repoduce test to demonstrate this im not convinced we should make this change.","commit_id":"30e57c9e9b0fdab02afbda1ccb116dcebf5d4abb"}],"/PATCHSET_LEVEL":[{"author":{"_account_id":9816,"name":"Takashi Kajinami","email":"kajinamit@oss.nttdata.com","username":"kajinamit"},"change_message_id":"f734cc9c9238fffb89296550f12a572579dce4d0","unresolved":true,"context_lines":[],"source_content_type":"","patch_set":2,"id":"bb685d2b_d5736296","updated":"2024-11-10 12:47:48.000000000","message":"Please consider adding clear explanation to the commit message so that the other people can understand what is \"improved\" by this change.","commit_id":"a4519a96265f0eb6b947e558739dbfe1b3cdbc64"},{"author":{"_account_id":37436,"name":"Si Snow","display_name":"silh","email":"silh@chinatelecom.cn","username":"silh"},"change_message_id":"1db3d80a15f2b5c4743650314bc54643e26c83d7","unresolved":false,"context_lines":[],"source_content_type":"","patch_set":2,"id":"05d50573_f81fdbd8","updated":"2024-11-07 01:52:47.000000000","message":"recheck - failed test nova-grenade-multinode","commit_id":"a4519a96265f0eb6b947e558739dbfe1b3cdbc64"},{"author":{"_account_id":37436,"name":"Si Snow","display_name":"silh","email":"silh@chinatelecom.cn","username":"silh"},"change_message_id":"b695460f31dfb3a8e582c637e245ab42be0dc943","unresolved":false,"context_lines":[],"source_content_type":"","patch_set":2,"id":"16169774_5aaed46a","in_reply_to":"bb685d2b_d5736296","updated":"2024-11-11 03:35:13.000000000","message":"Done","commit_id":"a4519a96265f0eb6b947e558739dbfe1b3cdbc64"},{"author":{"_account_id":9816,"name":"Takashi Kajinami","email":"kajinamit@oss.nttdata.com","username":"kajinamit"},"change_message_id":"18c136b21a380e701b149c1a6e51eaf9c81424b8","unresolved":true,"context_lines":[],"source_content_type":"","patch_set":3,"id":"f7801666_fa7264ea","updated":"2024-11-11 02:02:08.000000000","message":"Hmm... IIUC you are now making BarbicanKeyManager concurrency-safe by https://review.opendev.org/c/openstack/castellan/+/934421 . What is the point of making additional fix in nova side ?","commit_id":"30e57c9e9b0fdab02afbda1ccb116dcebf5d4abb"},{"author":{"_account_id":37436,"name":"Si Snow","display_name":"silh","email":"silh@chinatelecom.cn","username":"silh"},"change_message_id":"b257bcb43274802e53ded73b6957dd1564a2a158","unresolved":false,"context_lines":[],"source_content_type":"","patch_set":3,"id":"3e9c7b6f_38e1d7cf","in_reply_to":"f7801666_fa7264ea","updated":"2024-11-11 03:36:29.000000000","message":"At the beginning, I considered circumventing this problem by reinstantiating the key manager on every call by  https://review.opendev.org/c/openstack/cinder/+/934223 and https://review.opendev.org/c/openstack/nova/+/934225.\n\nIn https://review.opendev.org/c/openstack/cinder/+/934223, Eric Harney suggested solving this problem completely by adding a lock in castellan.\n\nThis commit can be closed if it is not needed.","commit_id":"30e57c9e9b0fdab02afbda1ccb116dcebf5d4abb"},{"author":{"_account_id":9816,"name":"Takashi Kajinami","email":"kajinamit@oss.nttdata.com","username":"kajinamit"},"change_message_id":"3c8ae6e63353c8729fcb488d25e1e654801f849f","unresolved":false,"context_lines":[],"source_content_type":"","patch_set":4,"id":"f9650969_1d9f3cac","updated":"2025-07-10 15:04:46.000000000","message":"The problem is caused by the incorrect cachming mechanism in castellan which was fixed by https://review.opendev.org/c/openstack/castellan/+/952669 so we don\u0027t have to update the service code calling castellan.","commit_id":"c09ae18caf33a9d92f7316f9e77a97b6c18cf8ec"},{"author":{"_account_id":9816,"name":"Takashi Kajinami","email":"kajinamit@oss.nttdata.com","username":"kajinamit"},"change_message_id":"9ddf973a10a4e472a10d5e1ed178c5333f8fb4d1","unresolved":true,"context_lines":[],"source_content_type":"","patch_set":4,"id":"3ab931e7_0c433c80","in_reply_to":"89c35a46_93e581fd","updated":"2025-07-10 15:17:00.000000000","message":"Yeah we can backport it, maybe after 2025.2 GA to catch any problems before we start the backport process (unless the reporter or anyone else is eager to have it in stable branch and test it :-) )","commit_id":"c09ae18caf33a9d92f7316f9e77a97b6c18cf8ec"},{"author":{"_account_id":11604,"name":"sean mooney","email":"smooney@redhat.com","username":"sean-k-mooney"},"change_message_id":"48f11700c760e6ea96762f37a1a44461eceee92a","unresolved":true,"context_lines":[],"source_content_type":"","patch_set":4,"id":"89c35a46_93e581fd","in_reply_to":"f9650969_1d9f3cac","updated":"2025-07-10 15:14:15.000000000","message":"ack is there a plan to backport that fix in castellan?\nif so i think i agree that this shoudl be fix via the libary rather then the service projects.","commit_id":"c09ae18caf33a9d92f7316f9e77a97b6c18cf8ec"}],"nova/compute/api.py":[{"author":{"_account_id":11604,"name":"sean mooney","email":"smooney@redhat.com","username":"sean-k-mooney"},"change_message_id":"a0b86e5aca937b3c3969b0b21af96c1c3c0b7315","unresolved":true,"context_lines":[{"line_number":406,"context_line":"        self.host_api \u003d HostAPI(self.compute_rpcapi, self.servicegroup_api)"},{"line_number":407,"context_line":"        self.notifier \u003d rpc.get_notifier(\u0027compute\u0027)"},{"line_number":408,"context_line":"        if CONF.ephemeral_storage_encryption.enabled:"},{"line_number":409,"context_line":"            self.key_manager \u003d key_manager.API()"},{"line_number":410,"context_line":"        # Help us to record host in EventReporter"},{"line_number":411,"context_line":"        self.host \u003d CONF.host"},{"line_number":412,"context_line":""}],"source_content_type":"text/x-python","patch_set":3,"id":"8f05c1f9_dcb15cb5","side":"PARENT","line":409,"updated":"2024-11-11 03:07:31.000000000","message":"Why are you moving the key manager from the compute API object to _populate_instance_for_create\n\nwe have other functionalities that will use the key manager.","commit_id":"0a59078935b753ba30518915e15189e63c0dfc66"},{"author":{"_account_id":37436,"name":"Si Snow","display_name":"silh","email":"silh@chinatelecom.cn","username":"silh"},"change_message_id":"8df74231f643df4298e0ea1f03d597c08548b6d5","unresolved":false,"context_lines":[{"line_number":406,"context_line":"        self.host_api \u003d HostAPI(self.compute_rpcapi, self.servicegroup_api)"},{"line_number":407,"context_line":"        self.notifier \u003d rpc.get_notifier(\u0027compute\u0027)"},{"line_number":408,"context_line":"        if CONF.ephemeral_storage_encryption.enabled:"},{"line_number":409,"context_line":"            self.key_manager \u003d key_manager.API()"},{"line_number":410,"context_line":"        # Help us to record host in EventReporter"},{"line_number":411,"context_line":"        self.host \u003d CONF.host"},{"line_number":412,"context_line":""}],"source_content_type":"text/x-python","patch_set":3,"id":"3773f95c_72010e05","side":"PARENT","line":409,"in_reply_to":"8f05c1f9_dcb15cb5","updated":"2024-11-11 03:40:58.000000000","message":"According to https://bugs.launchpad.net/cinder/+bug/2086765, BarbicanKeyManager is not concurrency-safe.\nIn this change, I circumvented this problem by re-instantiating the key manager on every call","commit_id":"0a59078935b753ba30518915e15189e63c0dfc66"}]}