)]}'
{"/COMMIT_MSG":[{"author":{"_account_id":15343,"name":"Tim Burke","email":"tburke@nvidia.com","username":"tburke"},"change_message_id":"d33ade328f846e984032e32c931ec5a5ebf7f0a8","unresolved":false,"context_lines":[{"line_number":6,"context_line":""},{"line_number":7,"context_line":"Remove TODO re taking lock before deleting retiring db"},{"line_number":8,"context_line":""},{"line_number":9,"context_line":"Remove this TODO since we do not plan to take a lock before deleting a"},{"line_number":10,"context_line":"retiring db.  A lock might be required if there was a risk of the"},{"line_number":11,"context_line":"retiring db being modified between the cleaving context being checked"},{"line_number":12,"context_line":"and the db being deleted in _complete_sharding(). The following steps"},{"line_number":13,"context_line":"have been taken to avoid any such modification:"}],"source_content_type":"text/x-gerrit-commit-message","patch_set":1,"id":"9f6a8fd7_a3f5dc58","line":10,"range":{"start_line":9,"start_character":60,"end_line":10,"end_character":11},"updated":"2018-04-25 19:33:05.000000000","message":"This was the only reason we were even considering it, right? We trust sqlite to do appropriate locking for our changes to the fresh db?","commit_id":"eccc552f52ce3b3afebe9f58d1eaf243b4c5e477"},{"author":{"_account_id":15343,"name":"Tim Burke","email":"tburke@nvidia.com","username":"tburke"},"change_message_id":"d33ade328f846e984032e32c931ec5a5ebf7f0a8","unresolved":false,"context_lines":[{"line_number":13,"context_line":"have been taken to avoid any such modification:"},{"line_number":14,"context_line":""},{"line_number":15,"context_line":"1. No object updates are committed to the retiring db from the pending"},{"line_number":16,"context_line":"updates file [1]."},{"line_number":17,"context_line":""},{"line_number":18,"context_line":"2. No objects are replicated to the retiring db (nor the fresh db)"},{"line_number":19,"context_line":"after sharding begins [2]."}],"source_content_type":"text/x-gerrit-commit-message","patch_set":1,"id":"9f6a8fd7_a67baa89","line":16,"updated":"2018-04-25 19:33:05.000000000","message":"...and we\u0027re already confident that any metadata or shard range updates that might have been applied to the retiring db would already have been applied to the fresh db... right?\n\nNow I\u0027m a little worried about a POST coming in between us fetching metadata from the old broker and renaming the preshard broker into place... but, hey, replication will probably save us, right??","commit_id":"eccc552f52ce3b3afebe9f58d1eaf243b4c5e477"}]}
