)]}'
{"trove/instance/models.py":[{"author":{"_account_id":694,"name":"Tim Simpson","email":"tim.simpson@rackspace.com","username":"tim-simpson"},"change_message_id":"202db0fef8b20f8f9b7400e69e57600e5dc44834","unresolved":false,"context_lines":[{"line_number":170,"context_line":""},{"line_number":171,"context_line":"    @property"},{"line_number":172,"context_line":"    def is_running(self):"},{"line_number":173,"context_line":"        \"\"\"True if the service status indicates MySQL is up and running.\"\"\""},{"line_number":174,"context_line":"        return self.service_status.status in INSTANCE_RESPONSIVE_STATUSES"},{"line_number":175,"context_line":""},{"line_number":176,"context_line":"    @property"}],"source_content_type":"text/x-python","patch_set":1,"id":"AAAAUH%2F%2FIy8%3D","line":173,"updated":"2014-01-17 04:11:22.000000000","message":"Nitpick- the pydoc still says MySQL.","commit_id":"f7ef74b733a5a32ef028e363aa2517f8502279de"},{"author":{"_account_id":694,"name":"Tim Simpson","email":"tim.simpson@rackspace.com","username":"tim-simpson"},"change_message_id":"202db0fef8b20f8f9b7400e69e57600e5dc44834","unresolved":false,"context_lines":[{"line_number":187,"context_line":"        if self.db_info.task_status.is_error:"},{"line_number":188,"context_line":"            return InstanceStatus.ERROR"},{"line_number":189,"context_line":"        if InstanceStatus.ERROR \u003d\u003d self.db_info.server_status:"},{"line_number":190,"context_line":"            return InstanceStatus.ERROR"},{"line_number":191,"context_line":"        ### Check for taskmanager status."},{"line_number":192,"context_line":"        ACTION \u003d self.db_info.task_status.action"},{"line_number":193,"context_line":"        if InstanceTasks.BUILDING.action \u003d\u003d ACTION:"}],"source_content_type":"text/x-python","patch_set":1,"id":"AAAAUH%2F%2FIr0%3D","line":190,"updated":"2014-01-17 04:11:22.000000000","message":"I disagree with this change. First, there is a small chance it could lead to an API change, but more importantly I think its the wrong way philosophically.\n\nIn Trove we manage a logical Trove instance on the user\u0027s behalf that is actually made up of a number of other resources. So if we\u0027re building the instance, resizing it, or whatever, we return the task status while we let the logic in task manager do various things to the resources manipulated by Trove. Meanwhile, those resources may themselves change statuses multiple times (particularly the guest status). The task manager code may not be pretty, but I will vouch for the fact that we spend a long time thinking about it and what statuses we want the user to see.\n\nAnyway, during the time we\u0027re doing tasks, we control the status. If it should be set to error we can do that. I think setting it here is the wrong choice.","commit_id":"f7ef74b733a5a32ef028e363aa2517f8502279de"}]}