add kernel_info_timeout traitlet to wait for slow kernel startups

mpacer · mpacer · commit 56055fbd63e8 · 2018-06-06T12:38:05.000-07:00
This affects both the MappingKernelManager and the ZMQChannelsHandler(by extension). This allows one setting to apply to both startup andrestarting.
diff --git a/notebook/services/kernels/handlers.py b/notebook/services/kernels/handlers.py
@@ -104,7 +104,8 @@ class ZMQChannelsHandler(AuthenticatedZMQStreamHandler):
 
     @property
     def kernel_info_timeout(self):
-        return self.settings.get('kernel_info_timeout', 10)
+        km_default = self.kernel_manager.kernel_info_timeout
+        return self.settings.get('kernel_info_timeout', km_default)
 
     @property
     def iopub_msg_rate_limit(self):
diff --git a/notebook/services/kernels/kernelmanager.py b/notebook/services/kernels/kernelmanager.py
@@ -19,7 +19,7 @@
 from jupyter_client.session import Session
 from jupyter_client.multikernelmanager import MultiKernelManager
 from traitlets import (Any, Bool, Dict, List, Unicode, TraitError, Integer,
-       Instance, default, validate
+       Float, Instance, default, validate
 )
 
 from notebook.utils import to_os_path, exists
@@ -93,6 +93,22 @@ def _update_root_dir(self, proposal):
         no frontends are connected.
         """
     )
+    
+    kernel_info_timeout = Float(60, config=True,
+        help="""Timeout for giving up on a kernel (in seconds).
+
+        On starting and restarting kernels, we check whether the
+        kernel is running and responsive by sending kernel_info_requests.
+        This sets the timeout in seconds for how long the kernel can take
+        before being presumed dead. 
+        This affects the MappingKernelManager (which handles kernel restarts) 
+        and the ZMQChannelsHandler (which handles the startup).
+        """
+    )
+
+    kernel_info_timeout = Integer(30, config=True,
+        help="""This changes the amount of time that we wait for the kernel_info_reply the kernel restart is requested."""
+    )
 
     _kernel_buffers = Any()
     @default('_kernel_buffers')
@@ -305,7 +321,7 @@ def on_restart_failed():
         kernel.session.send(channel, "kernel_info_request")
         channel.on_recv(on_reply)
         loop = IOLoop.current()
-        timeout = loop.add_timeout(loop.time() + 30, on_timeout)
+        timeout = loop.add_timeout(loop.time() + self.kernel_info_timeout, on_timeout)
         return future
 
     def notify_connect(self, kernel_id):
@@ -434,4 +450,3 @@ def cull_kernel_if_idle(self, kernel_id):
                 self.log.warning("Culling '%s' kernel '%s' (%s) with %d connections due to %s seconds of inactivity.",
                                  kernel.execution_state, kernel.kernel_name, kernel_id, connections, idle_duration)
                 self.shutdown_kernel(kernel_id)
-
