ctlmgr: support immediate controller retry

artiq/frontend/artiq_ctlmgr.py

@@ -1,6 +1,7 @@
 #!/usr/bin/env python3
 
 import asyncio
+import atexit
 import argparse
 import os
 import logging
@@ -8,9 +9,9 @@ import shlex
 import socket
 
 from artiq.protocols.sync_struct import Subscriber
-from artiq.protocols.pc_rpc import AsyncioClient
+from artiq.protocols.pc_rpc import AsyncioClient, Server
 from artiq.tools import verbosity_args, init_logger
-from artiq.tools import asyncio_process_wait_timeout
+from artiq.tools import TaskObject, asyncio_process_wait_timeout, Condition
 
 
 logger = logging.getLogger(__name__)
@@ -29,8 +30,11 @@ def get_argparser():
         "--retry-master", default=5.0, type=float,
         help="retry timer for reconnecting to master")
     parser.add_argument(
-        "--retry-command", default=5.0, type=float,
-        help="retry timer for restarting a controller command")
+        "--bind", default="::1",
+        help="hostname or IP address to bind to")
+    parser.add_argument(
+        "--bind-port", default=3249, type=int,
+        help="TCP port to listen to for control (default: %(default)d)")
     return parser
 
 
@@ -48,6 +52,7 @@ class Controller:
         self.term_timeout = ddb_entry.get("term_timeout", 30)
 
         self.retry_timer_cur = self.retry_timer
+        self.retry_now = Condition()
         self.process = None
         self.launch_task = asyncio.Task(self.launcher())
 
@@ -109,8 +114,13 @@ class Controller:
                     logger.warning("Controller %s failed to start", self.name)
                 else:
                     logger.warning("Controller %s exited", self.name)
-                logger.warning("Restarting in %.1f seconds", self.retry_timer_cur)
-                yield from asyncio.sleep(self.retry_timer_cur)
+                logger.warning("Restarting in %.1f seconds",
+                               self.retry_timer_cur)
+                try:
+                    yield from asyncio.wait_for(self.retry_now.wait(),
+                                                self.retry_timer_cur)
+                except asyncio.TimeoutError:
+                    pass
                 self.retry_timer_cur *= self.retry_timer_backoff
         except asyncio.CancelledError:
             yield from self._terminate()
@@ -208,18 +218,26 @@ class ControllerDB:
         return self.current_controllers
 
 
+class ControllerManager(TaskObject):
+    def __init__(self, server, port, retry_master):
+        self.server = server
+        self.port = port
+        self.retry_master = retry_master
+        self.controller_db = ControllerDB()
+
     @asyncio.coroutine
-def ctlmgr(server, port, retry_master):
-    controller_db = ControllerDB()
+    def _do(self):
         try:
-        subscriber = Subscriber("devices", controller_db.sync_struct_init)
+            subscriber = Subscriber("devices",
+                                    self.controller_db.sync_struct_init)
             while True:
                 try:
                     def set_host_filter():
                         s = subscriber.writer.get_extra_info("socket")
                         localhost = s.getsockname()[0]
-                    controller_db.set_host_filter(localhost)
-                yield from subscriber.connect(server, port, set_host_filter)
+                        self.controller_db.set_host_filter(localhost)
+                    yield from subscriber.connect(self.server, self.port,
+                                                  set_host_filter)
                     try:
                         yield from asyncio.wait_for(subscriber.receive_task, None)
                     finally:
@@ -230,12 +248,17 @@ def ctlmgr(server, port, retry_master):
                         e.__class__.__name__, str(e))
                 else:
                     logger.warning("Connection to master lost")
-            logger.warning("Retrying in %.1f seconds", retry_master)
-            yield from asyncio.sleep(retry_master)
+                logger.warning("Retrying in %.1f seconds", self.retry_master)
+                yield from asyncio.sleep(self.retry_master)
         except asyncio.CancelledError:
             pass
         finally:
-        yield from controller_db.current_controllers.shutdown()
+            yield from self.controller_db.current_controllers.shutdown()
+
+    def retry_now(self, k):
+        """If a controller is disabled and pending retry, perform that retry
+        now."""
+        self.controller_db.current_controllers.active[k].retry_now.notify()
 
 
 def main():
@@ -247,18 +270,22 @@ def main():
         asyncio.set_event_loop(loop)
     else:
         loop = asyncio.get_event_loop()
+    atexit.register(lambda: loop.close())
 
-    try:
-        task = asyncio.Task(ctlmgr(
-            args.server, args.port, args.retry_master))
-        try:
-            loop.run_forever()
-        finally:
-            task.cancel()
-            loop.run_until_complete(asyncio.wait_for(task, None))
+    ctlmgr = ControllerManager(args.server, args.port, args.retry_master)
+    ctlmgr.start()
+    atexit.register(lambda: loop.run_until_complete(ctlmgr.stop()))
+
+    class CtlMgrRPC:
+        retry_now = ctlmgr.retry_now
+
+    rpc_target = CtlMgrRPC()
+    rpc_server = Server({"ctlmgr": rpc_target}, builtin_terminate=True)
+    loop.run_until_complete(rpc_server.start(args.bind, args.bind_port))
+    atexit.register(lambda: loop.run_until_complete(rpc_server.stop()))
+
+    loop.run_until_complete(rpc_server.wait_terminate())
 
-    finally:
-        loop.close()
 
 if __name__ == "__main__":
     main()

doc/manual/default_network_ports.rst

@@ -8,6 +8,8 @@ Default network ports
 +--------------------------+--------------+
 | Core device (mon/inj)    | 3250 (UDP)   |
 +--------------------------+--------------+
+| Controller manager       | 3249         |
++--------------------------+--------------+
 | Master (notifications)   | 3250         |
 +--------------------------+--------------+
 | Master (control)         | 3251         |