Handle the exception for running the profiler 02/67502/2
authordongwenjuan <dongwenjuan1122@gmail.com>
Tue, 9 Apr 2019 02:22:52 +0000 (10:22 +0800)
committerdongwenjuan <dongwenjuan1122@gmail.com>
Tue, 9 Apr 2019 06:16:21 +0000 (14:16 +0800)
Congress as inspector, currently we can't get the vmdown time
and hostdown time unless we get the log of congress and extract the times.

Change-Id: Ie8b9dff6048040a1f834033c4873c168225e9d8c
Signed-off-by: dongwenjuan <dongwenjuan1122@gmail.com>
doctor_tests/scenario/fault_management.py

index a110b88..5216aa8 100644 (file)
@@ -209,6 +209,10 @@ class FaultManagement(object):
         detected = self.monitor.detected_time
         notified = self.consumer.notified_time
 
+        if None in [vmdown, hostdown, detected, notified]:
+            self.log.info('one of the time for profiler is None, return')
+            return
+
         # TODO(yujunz) check the actual delay to verify time sync status
         # expected ~1s delay from $trigger to $linkdown
         relative_start = linkdown