)]}'
{"/COMMIT_MSG":[{"author":{"_account_id":9531,"name":"liuyulong","display_name":"LIU Yulong","email":"i@liuyulong.me","username":"LIU-Yulong"},"change_message_id":"542d3e7046e1794b010b12bdbbc162b446f1740c","unresolved":false,"context_lines":[{"line_number":6,"context_line":""},{"line_number":7,"context_line":"Add a random initial_delay when agent starts"},{"line_number":8,"context_line":""},{"line_number":9,"context_line":"When managing bigger infrastructure, we may restart the neutron agents"},{"line_number":10,"context_line":"from one place and all together."},{"line_number":11,"context_line":"As a result, the agents are reporting to neutron server at the same"},{"line_number":12,"context_line":"time, which can lead to some load on server side."},{"line_number":13,"context_line":"By adding a random delay before reporting the first time, it will allow"}],"source_content_type":"text/x-gerrit-commit-message","patch_set":2,"id":"7faddb67_1a3d712a","line":10,"range":{"start_line":9,"start_character":37,"end_line":10,"end_character":32},"updated":"2019-07-15 09:17:54.000000000","message":"IMO, it is not a good practice to do so. For my experience, how many neutron RPC workers you have, how many agents you can restart parallelly. It is a predictable number. Then your restart action will be safe IMO.","commit_id":"3ebdeb12d7682cfcec95e2c21399423de3201e80"}],"neutron/plugins/ml2/drivers/openvswitch/agent/ovs_neutron_agent.py":[{"author":{"_account_id":9531,"name":"liuyulong","display_name":"LIU Yulong","email":"i@liuyulong.me","username":"LIU-Yulong"},"change_message_id":"542d3e7046e1794b010b12bdbbc162b446f1740c","unresolved":false,"context_lines":[{"line_number":322,"context_line":"        report_interval \u003d agent_conf.report_interval"},{"line_number":323,"context_line":"        if report_interval:"},{"line_number":324,"context_line":"            # add random initial delay to allow agents to check in smoothly"},{"line_number":325,"context_line":"            initial_delay \u003d random.randint(0, report_interval)"},{"line_number":326,"context_line":"            heartbeat \u003d loopingcall.FixedIntervalLoopingCall("},{"line_number":327,"context_line":"                self._report_state)"},{"line_number":328,"context_line":"            heartbeat.start(interval\u003dreport_interval,"}],"source_content_type":"text/x-python","patch_set":2,"id":"7faddb67_fa2895e2","line":325,"range":{"start_line":325,"start_character":46,"end_line":325,"end_character":61},"updated":"2019-07-15 09:17:54.000000000","message":"So, if this is set to 300s to 600s, that means some agent will be available after this long time. This is not a good approach IMO.","commit_id":"3ebdeb12d7682cfcec95e2c21399423de3201e80"}]}