)]}'
{"neutron/db/agents_db.py":[{"author":{"_account_id":8873,"name":"Assaf Muller","email":"amuller@redhat.com","username":"amuller"},"change_message_id":"fabdb760a6740053af0eca06bc26077bbba08401","unresolved":false,"context_lines":[{"line_number":349,"context_line":"                res[\u0027heartbeat_timestamp\u0027] \u003d current_time"},{"line_number":350,"context_line":"                if agent_state.get(\u0027start_flag\u0027):"},{"line_number":351,"context_line":"                    res[\u0027started_at\u0027] \u003d current_time"},{"line_number":352,"context_line":"                greenthread.sleep(0)"},{"line_number":353,"context_line":"                self._log_heartbeat(agent_state, agent_db, configurations_dict)"},{"line_number":354,"context_line":"                agent_db.update(res)"},{"line_number":355,"context_line":"            except ext_agent.AgentNotFoundByTypeHost:"}],"source_content_type":"text/x-python","patch_set":1,"id":"fa0719c6_4990b843","line":352,"updated":"2016-03-23 00:41:16.000000000","message":"What do you think about these sleeps after we moved to the new mysql driver? Would they contribute to deadlocks?","commit_id":"1f22027c98fdf06064d43a6ee30042adddf867d1"},{"author":{"_account_id":6072,"name":"Eugene Nikanorov","email":"enikanorov@mirantis.com","username":"enikanorov"},"change_message_id":"2c7b06b78f41e4febacbdb2219353da85ed4f17c","unresolved":false,"context_lines":[{"line_number":349,"context_line":"                res[\u0027heartbeat_timestamp\u0027] \u003d current_time"},{"line_number":350,"context_line":"                if agent_state.get(\u0027start_flag\u0027):"},{"line_number":351,"context_line":"                    res[\u0027started_at\u0027] \u003d current_time"},{"line_number":352,"context_line":"                greenthread.sleep(0)"},{"line_number":353,"context_line":"                self._log_heartbeat(agent_state, agent_db, configurations_dict)"},{"line_number":354,"context_line":"                agent_db.update(res)"},{"line_number":355,"context_line":"            except ext_agent.AgentNotFoundByTypeHost:"}],"source_content_type":"text/x-python","patch_set":1,"id":"fa0719c6_c9d20800","line":352,"in_reply_to":"fa0719c6_4990b843","updated":"2016-03-23 00:46:12.000000000","message":"I think the deadlock we get here is purely a write set certification failures which are caused by concurrent transactions being committed (remember i\u0027m addressing mutlimaster use case).\nIt\u0027s not something we used to have with previous sql driver and async operations under transactions.","commit_id":"1f22027c98fdf06064d43a6ee30042adddf867d1"}]}
