Make executor threads more robust

See bug #1189711

Prevent the executor thread getting killed by exceptions.

Change-Id: I0787fa619a6a8a35a303627f8ce119c93f2c8765

openstack-common.conf

@@ -1,6 +1,7 @@
 [DEFAULT]
 
 # The list of modules to copy from oslo-incubator.git
+module=excutils
 module=fixture.moxstubout
 module=jsonutils
 module=timeutils

oslo/messaging/_executors/impl_eventlet.py

@@ -20,6 +20,7 @@ import greenlet
 from oslo.config import cfg
 
 from oslo.messaging._executors import base
+from oslo.messaging.openstack.common import excutils
 
 _eventlet_opts = [
     cfg.IntOpt('rpc_thread_pool_size',
@@ -49,6 +50,7 @@ class EventletExecutor(base.ExecutorBase):
         if self._thread is not None:
             return
 
+        @excutils.forever_retry_uncaught_exceptions
         def _executor_thread():
             try:
                 while True:

oslo/messaging/openstack/common/excutils.py

@@ -0,0 +1,98 @@
+# vim: tabstop=4 shiftwidth=4 softtabstop=4
+
+# Copyright 2011 OpenStack Foundation.
+# Copyright 2012, Red Hat, Inc.
+#
+#    Licensed under the Apache License, Version 2.0 (the "License"); you may
+#    not use this file except in compliance with the License. You may obtain
+#    a copy of the License at
+#
+#         http://www.apache.org/licenses/LICENSE-2.0
+#
+#    Unless required by applicable law or agreed to in writing, software
+#    distributed under the License is distributed on an "AS IS" BASIS, WITHOUT
+#    WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. See the
+#    License for the specific language governing permissions and limitations
+#    under the License.
+
+"""
+Exception related utilities.
+"""
+
+import logging
+import sys
+import time
+import traceback
+
+from oslo.messaging.openstack.common.gettextutils import _  # noqa
+
+
+class save_and_reraise_exception(object):
+    """Save current exception, run some code and then re-raise.
+
+    In some cases the exception context can be cleared, resulting in None
+    being attempted to be re-raised after an exception handler is run. This
+    can happen when eventlet switches greenthreads or when running an
+    exception handler, code raises and catches an exception. In both
+    cases the exception context will be cleared.
+
+    To work around this, we save the exception state, run handler code, and
+    then re-raise the original exception. If another exception occurs, the
+    saved exception is logged and the new exception is re-raised.
+
+    In some cases the caller may not want to re-raise the exception, and
+    for those circumstances this context provides a reraise flag that
+    can be used to suppress the exception.  For example:
+
+    except Exception:
+        with save_and_reraise_exception() as ctxt:
+            decide_if_need_reraise()
+            if not should_be_reraised:
+                ctxt.reraise = False
+    """
+    def __init__(self):
+        self.reraise = True
+
+    def __enter__(self):
+        self.type_, self.value, self.tb, = sys.exc_info()
+        return self
+
+    def __exit__(self, exc_type, exc_val, exc_tb):
+        if exc_type is not None:
+            logging.error(_('Original exception being dropped: %s'),
+                          traceback.format_exception(self.type_,
+                                                     self.value,
+                                                     self.tb))
+            return False
+        if self.reraise:
+            raise self.type_, self.value, self.tb
+
+
+def forever_retry_uncaught_exceptions(infunc):
+    def inner_func(*args, **kwargs):
+        last_log_time = 0
+        last_exc_message = None
+        exc_count = 0
+        while True:
+            try:
+                return infunc(*args, **kwargs)
+            except Exception as exc:
+                if exc.message == last_exc_message:
+                    exc_count += 1
+                else:
+                    exc_count = 1
+                # Do not log any more frequently than once a minute unless
+                # the exception message changes
+                cur_time = int(time.time())
+                if (cur_time - last_log_time > 60 or
+                        exc.message != last_exc_message):
+                    logging.exception(
+                        _('Unexpected exception occurred %d time(s)... '
+                          'retrying.') % exc_count)
+                    last_log_time = cur_time
+                    last_exc_message = exc.message
+                    exc_count = 0
+                # This should be a very rare event. In case it isn't, do
+                # a sleep.
+                time.sleep(1)
+    return inner_func