diff --git a/SAS/TMSS/backend/services/scheduling/lib/constraints.py b/SAS/TMSS/backend/services/scheduling/lib/constraints.py
index 2857b7193d07c5486d207614623c0e53f55dcd11..32cde979c0c8773e0305ca7ff60f73c47b73d847 100644
--- a/SAS/TMSS/backend/services/scheduling/lib/constraints.py
+++ b/SAS/TMSS/backend/services/scheduling/lib/constraints.py
@@ -1779,7 +1779,20 @@ def get_earliest_possible_start_time_taking_used_stations_into_account(schedulin
latest_blocking_stop_time = max([unit.on_sky_stop_time for unit in blocking_units])
from .dynamic_scheduling import DEFAULT_INTER_OBSERVATION_GAP
advanced_lower_bound = latest_blocking_stop_time + DEFAULT_INTER_OBSERVATION_GAP
- return get_earliest_possible_start_time_taking_used_stations_into_account(scheduling_unit, advanced_lower_bound, upper_bound, gridder, raise_if_interruped)
+ if advanced_lower_bound < upper_bound:
+ logger.debug("get_earliest_possible_start_time_taking_used_stations_into_account: SUB id=%s earliest_possible_start_time='%s' is blocked by units %s. Advancing window from lower_bound='%s' to new lower_bound='%s', upper_bound='%s'",
+ scheduling_unit.id, earliest_possible_start_time, ','.join(str(u.id) for u in blocking_units), lower_bound, advanced_lower_bound, upper_bound)
+
+ advanced_earliest_possible_start_time = get_earliest_possible_start_time_taking_used_stations_into_account(scheduling_unit, advanced_lower_bound, upper_bound, gridder, raise_if_interruped)
+
+ if advanced_earliest_possible_start_time is not None:
+ logger.debug("get_earliest_possible_start_time_taking_used_stations_into_account: SUB id=%s advanced_earliest_possible_start_time='%s' lower_bound='%s', upper_bound='%s'",
+ scheduling_unit.id, advanced_earliest_possible_start_time, advanced_lower_bound, upper_bound)
+ return advanced_earliest_possible_start_time
+
+ logger.info("get_earliest_possible_start_time_taking_used_stations_into_account: SUB id=%s is blocked by units %s at earliest_possible_start_time='%s'",
+ scheduling_unit.id, ','.join(str(u.id) for u in blocking_units), earliest_possible_start_time)
+ return None
return earliest_possible_start_time
@@ -1969,7 +1982,7 @@ def get_weighted_start_time(scheduling_unit: models.SchedulingUnitBlueprint, low
logger.debug("get_weighted_start_time: SUB id=%s returning earliest='%s' - window=['%s', '%s']", scheduling_unit.id, earliest_possible_start_time, lower_bound, upper_bound)
return earliest_possible_start_time
- logger.warning("get_weighted_start_time: SUB id=%s could not compute weighted_start_time - window=['%s', '%s']", scheduling_unit.id, lower_bound, upper_bound)
+ logger.debug("get_weighted_start_time: SUB id=%s could not compute weighted_start_time - window=['%s', '%s']", scheduling_unit.id, lower_bound, upper_bound)
return None
@@ -1993,30 +2006,33 @@ def compute_individual_and_weighted_scores(scheduling_units: [models.SchedulingU
logger.info("compute_individual_and_weighted_scores: scored unit [%d/%d] %.1f%% %s",
i+1, len(scheduling_units), 100.0 * (i+1) / len(scheduling_units), scored_su)
else:
- logger.warning("compute_individual_and_weighted_scores: could not determine scores/start_time for unit [%d/%d] %.1f%% id=%d",
+ logger.info("compute_individual_and_weighted_scores: could not determine scores/start_time for unit [%d/%d] %.1f%% id=%d",
i+1, len(scheduling_units), 100.0 * (i+1) / len(scheduling_units), su.id)
except Exception as e:
logger.exception(e)
- # create/get a weight per unique score_name, caching it in a dict, thus reducing the number of be calls.
- weight_for_score_name = {}
- for score_name in sorted(list(score_names)):
- weight_factor, created = models.SchedulingConstraintsWeightFactor.objects.get_or_create(constraint_name=score_name,
- defaults={'weight': 1.0})
- weight_for_score_name[score_name] = weight_factor.weight
-
- logger.info('compute_individual_and_weighted_scores: weights: %s' , ' '.join(['%s=%.4f' % (score_name, weight_for_score_name[score_name]) for score_name in sorted(list(score_names))]))
-
- # do the actual weighting
- for i, scored_su in enumerate(scored_scheduling_units):
- # compute weighted total score
- weighted_score = 0.0
- for score_key, score_value in scored_su.scores.items():
- weight = weight_for_score_name[score_key]
- weighted_score += weight * score_value
- scored_su.weighted_score = weighted_score / float(len(scored_su.scores))
- logger.info("compute_individual_and_weighted_scores: weighted scores for unit [%d/%d] %.1f%% %s", i+1, len(scored_scheduling_units), 100.0 * (i + 1) / len(scored_scheduling_units), scored_su)
+ if scored_scheduling_units:
+ logger.info('compute_individual_and_weighted_scores: scored %d/%d units', len(scored_scheduling_units), len(scheduling_units))
+
+ # create/get a weight per unique score_name, caching it in a dict, thus reducing the number of be calls.
+ weight_for_score_name = {}
+ for score_name in sorted(list(score_names)):
+ weight_factor, created = models.SchedulingConstraintsWeightFactor.objects.get_or_create(constraint_name=score_name,
+ defaults={'weight': 1.0})
+ weight_for_score_name[score_name] = weight_factor.weight
+
+ logger.info('compute_individual_and_weighted_scores: applying weights: %s' , ' '.join(['%s=%.4f' % (score_name, weight_for_score_name[score_name]) for score_name in sorted(list(score_names))]))
+
+ # do the actual weighting
+ for i, scored_su in enumerate(scored_scheduling_units):
+ # compute weighted total score
+ weighted_score = 0.0
+ for score_key, score_value in scored_su.scores.items():
+ weight = weight_for_score_name[score_key]
+ weighted_score += weight * score_value
+ scored_su.weighted_score = weighted_score / float(len(scored_su.scores))
+ logger.info("compute_individual_and_weighted_scores: weighted scores for unit [%d/%d] %.1f%% %s", i+1, len(scored_scheduling_units), 100.0 * (i + 1) / len(scored_scheduling_units), scored_su)
return scored_scheduling_units
diff --git a/SAS/TMSS/backend/services/scheduling/lib/dynamic_scheduling.py b/SAS/TMSS/backend/services/scheduling/lib/dynamic_scheduling.py
index e3aa288490f4817a0ee5627fe597974924be1da7..5520e56f1100c6f9962d76bb2ce4d375cee8d7ec 100644
--- a/SAS/TMSS/backend/services/scheduling/lib/dynamic_scheduling.py
+++ b/SAS/TMSS/backend/services/scheduling/lib/dynamic_scheduling.py
@@ -554,6 +554,7 @@ class Scheduler:
# scanning from now+DEFAULT_NEXT_STARTTIME_GAP may include/overlap with already running and/or scheduled observations. That's ok. There are rules to handle the precedence who wins.
lower_bound_start_time = round_to_second_precision(max(datetime.utcnow() + DEFAULT_NEXT_STARTTIME_GAP,
min([su.earliest_possible_cycle_start_time for su in candidate_units])))
+ lower_bound_start_time = datetime(2023, 10, 6, 15)
# upper bound of search window is at least a 24h later, or up unit latest cycle end time
upper_bound_stop_time = round_to_second_precision(max(lower_bound_start_time + timedelta(days=1),
@@ -710,10 +711,12 @@ class Scheduler:
if (upper_bound_stop_time - lower_bound_start_time) <= 2*DEFAULT_INTER_OBSERVATION_GAP + timedelta(minutes=1):
continue # gap is too small
- logger.info("place_B_priority_units_in_gaps: evaluating %s B-queue units in %d[min]-wide gap ['%s', '%s') next to unit id=%s",
+ logger.info("place_B_priority_units_in_gaps: evaluating %s B-queue units in %d[min]-wide gap ['%s', '%s') %s unit id=%s",
schedulable_units_queue_B.count(),
(upper_bound_stop_time-lower_bound_start_time).total_seconds()/60,
- lower_bound_start_time, upper_bound_stop_time, scheduling_unit.id)
+ lower_bound_start_time, upper_bound_stop_time,
+ "before" if gap[1] <= scheduling_unit.scheduled_start_time else "after",
+ scheduling_unit.id)
best_B_candidate_for_gap = self.find_best_next_schedulable_unit(schedulable_units_queue_B, lower_bound_start_time=lower_bound_start_time, upper_bound_stop_time=upper_bound_stop_time)
if best_B_candidate_for_gap is not None and best_B_candidate_for_gap.scheduling_unit is not None: