Fixed "0" interval rules, that could hang UDS index page & consume all

memory

server/src/uds/REST/methods/calendarrules.py

@@ -109,6 +109,10 @@ class CalendarRules(DetailHandler):  # pylint: disable=too-many-public-methods
         # We need this fields for all
         logger.debug('Saving rule {0} / {1}'.format(parent, item))
         fields = self.readFieldsFromParams(['name', 'comments', 'frequency', 'start', 'end', 'interval', 'duration', 'duration_unit'])
+
+        if int(fields['interval']) == 0:
+            self.invalidItemException('Element can\'t have a 0 interval')
+
         # Convert timestamps to datetimes
         fields['start'] = datetime.datetime.fromtimestamp(fields['start'])
         if fields['end'] != None:

server/src/uds/core/util/calendar/__init__.py

@@ -46,7 +46,7 @@ import six
 import bitarray
 import logging
 
-__updated__ = '2016-04-26'
+__updated__ = '2016-10-31'
 
 
 logger = logging.getLogger(__name__)
@@ -85,7 +85,11 @@ class CalendarChecker(object):
             ruleDurationMinutes = rule.duration_as_minutes
             ruleFrequencyMinutes = rule.frequency_as_minutes
 
-            ruleDurationMinutes = ruleDurationMinutes
+            # Skip "bogus" definitions
+            if ruleDurationMinutes == 0 or ruleFrequencyMinutes == 0:
+                continue
+
+            # ruleDurationMinutes = ruleDurationMinutes
             # Relative start, rrule can "spawn" the days, so we get the start at least the ruleDurationMinutes of rule to see if it "matches"
             # This means, we need the previous matching day to be "executed" so we can get the "actives" correctly
             diff = ruleFrequencyMinutes if ruleFrequencyMinutes > ruleDurationMinutes else ruleDurationMinutes