I have migrated from Optaplanner version 7.5 to 7.10. While migration I saw bulk(Output) is not the same. While running for another time I am not getting the same output. Can you share which are necessary steps I need to consider while migration? Can you help me out?
The Job schedule is different too. Screenshot attached below:
The primary difference is for Run 1 executes LS 11040 but Run 2 for the same time gets to reach LS 11600 plus with environment and other run parameters maintained without being changed.
Job Schedule diffference
log1_run
11:35:06.421 [main ] INFO Solving started: time spent (225), best score (-1722init/0hard/0medium/0soft), environment mode (REPRODUCIBLE), random (JDK with seed 0).
11:35:06.682 [main ] DEBUG CH step (0), time spent (487), score (-1721init/0hard/0medium/-900000soft), selected move count (543), picked move (TimeWindowedCustomer-1 {null -> Vehicle-393}).
11:35:06.756 [main ] DEBUG CH step (1), time spent (561), score (-1720init/0hard/0medium/-1560000soft), selected move count (544), picked move (TimeWindowedCustomer-2 {null -> TimeWindowedCustomer-7}).
11:35:07.229 [main ] DEBUG CH step (2), time spent (1034), score (-1719init/0hard/0medium/-2520000soft), selected move count (545), picked move (TimeWindowedCustomer-3 {null -> Vehicle-396}).
11:35:07.266 [main ] DEBUG CH step (3), time spent (1071), score (-1718init/0hard/0medium/-3360000soft), selected move count (546), picked move (TimeWindowedCustomer-4 {null -> Vehicle-328}).
--------------------------------------------------
--------------------------------------------------
--------------------------------------------------
--------------------------------------------------
11:37:06.194 [main ] DEBUG LS step (11039), time spent (119999), score (0hard/-45187medium/-1556640000soft), best score (0hard/-45154medium/-1549560000soft), accepted/selected move count (1/3), picked move (TimeWindowedCustomer-1 {Vehicle-220} <-tailChainSwap-> TimeWindowedCustomer-8 {Vehicle-78}).
11:37:06.200 [main ] DEBUG LS step (11040), time spent (120005), score (-54512774442hard/-30651medium/-1659540000soft), best score (0hard/-45154medium/-1549560000soft), accepted/selected move count (0/1), picked move (TimeWindowedCustomer-12 {TimeWindowedCustomer-4} <-tailChainSwap-> null {TimeWindowedCustomer-16}).
11:37:06.203 [main ] INFO Local Search phase (1) ended: time spent (120008), best score (0hard/-45154medium/-1549560000soft), score calculation speed (5349/sec), step total (11041).
11:37:06.203 [main ] INFO Solving ended: time spent (120008), best score (0hard/-45154medium/-1549560000soft), score calculation speed (22645/sec), phase total (2), environment mode (REPRODUCIBLE).
log2 _run
11:38:29.684 [main ] INFO Solving started: time spent (213), best score (-1722init/0hard/0medium/0soft), environment mode (REPRODUCIBLE), random (JDK with seed 0).
11:38:29.955 [main ] DEBUG CH step (0), time spent (485), score (-1721init/0hard/0medium/-900000soft), selected move count (543), picked move (TimeWindowedCustomer-97041025 {null -> Vehicle-393}).
11:38:30.027 [main ] DEBUG CH step (1), time spent (557), score (-1720init/0hard/0medium/-1560000soft), selected move count (544), picked move (TimeWindowedCustomer-1 {null -> TimeWindowedCustomer-6}).
11:38:30.075 [main ] DEBUG CH step (2), time spent (605), score (-1719init/0hard/0medium/-2520000soft), selected move count (545), picked move (TimeWindowedCustomer-2 {null -> Vehicle-396}).
11:38:30.114 [main ] DEBUG CH step (3), time spent (644), score (-1718init/0hard/0medium/-3360000soft), selected move count (546), picked move (TimeWindowedCustomer-3 {null -> Vehicle-328}).
-------------------------------------------
--------------------------------------------
11:40:25.467 [main ] DEBUG LS step (11039), time spent (116004), score (0hard/-45187medium/-1556640000soft), best score (0hard/-45154medium/-1549560000soft), accepted/selected move count (1/3), picked move (TimeWindowedCustomer-1 {Vehicle-220} <-tailChainSwap-> TimeWindowedCustomer-8 {Vehicle-78}).
11:40:25.470 [main ] DEBUG LS step (11040), time spent (116011), score (0hard/-45187medium/-1556640000soft), best score (0hard/-45154medium/-1549560000soft), accepted/selected move count (0/1), picked move (TimeWindowedCustomer-12 {TimeWindowedCustomer-4} <-tailChainSwap-> null {TimeWindowedCustomer-16}).
--------------------------------------------
--------------------------------------------
11:40:29.474 [main ] DEBUG LS step (11674), time spent (120004), score (-4200029hard/-45093medium/-1554780000soft), best score (0hard/-45064medium/-1558380000soft), accepted/selected move count (0/1), picked move (TimeWindowedCustomer-9 {TimeWindowedCustomer-5 -> TimeWindowedCustomer-2}).
11:40:29.477 [main ] INFO Local Search phase (1) ended: time spent (120007), best score (0hard/-45064medium/-1558380000soft), score calculation speed (5451/sec), step total (11675).
11:40:29.477 [main ] INFO Solving ended: time spent (120007), best score (0hard/-45064medium/-1558380000soft), score calculation speed (22762/sec), phase total (2), environment mode (REPRODUCIBLE).
Reproducibility is only guaranteed on each step index, not on time (although they are correlated). When using time spent termination, for example 2 minutes, there is no guarantee how much CPU cycles OptaPlanner will get and how and when the JVM will hotspot optimize. In some situations, it might actually get almost no CPU cycles. In other situations it gets plenty - in any case OptaPlanner uses what it gets, nothing goes to waste.
Looking at the step index level, the two runs are exactly the same (except for the time spent):
First run:
11:37:06.194 ...LS step (11039), time spent (119999), score (0hard/-45187medium/-1556640000soft), best score (0hard/-45154medium/-1549560000soft), accepted/selected move count (1/3), picked move (TimeWindowedCustomer-1 {Vehicle-220} <-tailChainSwap-> TimeWindowedCustomer-8 {Vehicle-78}).
Second run:
11:40:25.467 ...LS step (11039), time spent (116004), score (0hard/-45187medium/-1556640000soft), best score (0hard/-45154medium/-1549560000soft), accepted/selected move count (1/3), picked move (TimeWindowedCustomer-1 {Vehicle-220} <-tailChainSwap-> TimeWindowedCustomer-8 {Vehicle-78}).
To get the exact same result between 2 results, use a step limit termination instead.
Related
I am working on optimizing the Construction Heuristics for the solution domain
I have a ValueRangerProvider as the following.
#ValueRangeProvider(id = "timeRange")
public CountableValueRange<LocalDateTime> timeRange() {
final LocalDateTime start = LocalDateTime.now().plusHours(1).truncatedTo(ChronoUnit.HOURS);
final LocalDateTime end = LocalDateTime.now().plusYears(1).truncatedTo(ChronoUnit.HOURS);
return Stream.iterate(start, d -> d.plusHours(1))
.takeWhile(localDateTime -> localDateTime.isBefore(end))
.collect(Collectors.toList());
}
This #ValueRangeProvider provides LocalDateTime instances for every hour between the start and end to be used in a Time Grain model. The timeRange value range is the only #PlannningVariable in the #PlanningEntity class.
I also have a following solverConfig.xml:
<?xml version="1.0" encoding="UTF-8"?>
<solver xmlns="https://www.optaplanner.org/xsd/solver" xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"
xsi:schemaLocation="https://www.optaplanner.org/xsd/solver/solver.xsd">
<scanAnnotatedClasses/>
<scoreDirectorFactory>
<constraintProviderClass>org.acme.optaplanner.solver.MyConstraintProvider</constraintProviderClass>
<initializingScoreTrend>ONLY_DOWN</initializingScoreTrend>
</scoreDirectorFactory>
<termination>
<bestScoreFeasible>true</bestScoreFeasible>
</termination>
<constructionHeuristic>
<constructionHeuristicType>FIRST_FIT</constructionHeuristicType>
<changeMoveSelector>
<selectionOrder>ORIGINAL</selectionOrder>
</changeMoveSelector>
<constructionHeuristic>
</solver>
During the Construction Heuristics, the objects initialization is more important than the quality of the solution. Given there is enough time slot that no hard constraint can occur.
I realized that the #ValueRangeProvider(id = "timeRange") values used by the previous steps are being re-used in the later steps during the Construction Heuristics.
1772 [main] DEBUG o.o.c.i.c.DefaultConstructionHeuristicPhase - CH step (0), time spent (157), score (-999init/0hard/-1soft), selected move count (1), picked move (Lesson{id=1, subject='Math I', pinned=false, startTime=null, durationInGrains=3} {null -> 2020-09-27T09:00}).
.
.
.
02:01:26.494 [main ] DEBUG CH step (497), time spent (14513), score (-502init/0hard/-5542soft), selected move count (664), picked move (Lesson{id=498, subject='Math I', pinned=false, startTime=null, durationInGrains=3} {null -> 2020-11-10T12:00}).
02:01:26.565 [main ] DEBUG CH step (498), time spent (14584), score (-501init/0hard/-5548soft), selected move count (666), picked move (Lesson{id=499, subject='Math I', pinned=false, startTime=null, durationInGrains=1} {null -> 2020-11-10T14:00}).
.
.
.
17456 [main] TRACE o.o.c.i.c.d.ConstructionHeuristicDecider - Move index (0), score (-500init/-2hard/-5571soft), move (Lesson{id=500, subject='Math I', pinned=false, startTime=null, durationInGrains=2} {null -> 2020-09-27T09:00}).
.
.
.
02:01:26.637 [main ] DEBUG CH step (499), time spent (14656), score (-500init/0hard/-5557soft), selected move count (667), picked move (Lesson{id=500, subject='Math I', pinned=false, startTime=null, durationInGrains=2} {null -> 2020-11-10T15:00}).
02:01:26.711 [main ] DEBUG CH step (500), time spent (14730), score (-499init/0hard/-5568soft), selected move count (668), picked move (Lesson{id=501, subject='Math I', pinned=false, startTime=null, durationInGrains=3} {null -> 2020-11-10T16:00}).
02:01:26.787 [main ] DEBUG CH step (501), time spent (14806), score (-498init/0hard/-5581soft), selected move count (670), picked move (Lesson{id=502, subject='Math I', pinned=false, startTime=null, durationInGrains=1} {null -> 2020-11-10T18:00}).
.
.
.
02:17:46.156 [main ] INFO Construction Heuristic phase (0) ended: time spent (98518), best score (0hard/-11165soft), score calculation speed (6783/sec), step total (1000).
02:17:46.165 [main ] INFO Solving ended: time spent (98527), best score (0hard/-11165soft), score calculation speed (6776/sec), phase total (2), environment mode (NON_REPRODUCIBLE).
The LocalDateTime of 2020-10-01T09:00 was used in the CH step(0) and CH step(499).
Since I have a hard constraint that constrains that two #PlanningEntity cannot use the same timeRange value I was wondering if the used timeRange value can be removed for future steps to use?
As an example;
I have changed the constructionHeuristic configuration
<constructionHeuristic>
<constructionHeuristicType>FIRST_FIT</constructionHeuristicType>
<changeMoveSelector>
<selectionOrder>RANDOM</selectionOrder>
</changeMoveSelector>
<forager>
<pickEarlyType>FIRST_FEASIBLE_SCORE_OR_NON_DETERIORATING_HARD</pickEarlyType>
</forager>
</constructionHeuristic>
Which lead to a much fast Construction Heuristics performance because the step was choosing #Planningvariables for #PlanningRangeProvider at random.
.
.
.
2535 [main] TRACE o.o.c.i.c.d.ConstructionHeuristicDecider - Move index (0), score (-2init/0hard/-3303soft), move (Lesson{id=998, subject='Math I', pinned=false, startTime=null, durationInGrains=2} {null -> 2020-10-13T20:00}).
2535 [main] DEBUG o.o.c.i.c.DefaultConstructionHeuristicPhase - CH step (997), time spent (1046), score (-2init/0hard/-3303soft), selected move count (1), picked move (Lesson{id=998, subject='Math I', pinned=false, startTime=null, durationInGrains=2} {null -> 2020-10-13T20:00}).
2536 [main] TRACE o.o.c.i.c.d.ConstructionHeuristicDecider - Move index (0), score (-1init/0hard/-3312soft), move (Lesson{id=999, subject='Math I', pinned=false, startTime=null, durationInGrains=3} {null -> 2021-08-26T22:00}).
2536 [main] DEBUG o.o.c.i.c.DefaultConstructionHeuristicPhase - CH step (998), time spent (1047), score (-1init/0hard/-3312soft), selected move count (1), picked move (Lesson{id=999, subject='Math I', pinned=false, startTime=null, durationInGrains=3} {null -> 2021-08-26T22:00}).
2537 [main] TRACE o.o.c.i.c.d.ConstructionHeuristicDecider - Move index (0), score (0hard/-3317soft), move (Lesson{id=1000, subject='Math I', pinned=false, startTime=null, durationInGrains=1} {null -> 2021-03-10T15:00}).
2537 [main] DEBUG o.o.c.i.c.DefaultConstructionHeuristicPhase - CH step (999), time spent (1048), score (0hard/-3317soft), selected move count (1), picked move (Lesson{id=1000, subject='Math I', pinned=false, startTime=null, durationInGrains=1} {null -> 2021-03-10T15:00}).
2540 [main] INFO o.o.c.i.c.DefaultConstructionHeuristicPhase - Construction Heuristic phase (0) ended: time spent (1051), best score (0hard/-3317soft), score calculation speed (1304/sec), step total (1000).
2549 [main] INFO o.o.core.impl.solver.DefaultSolver - Solving ended: time spent (1059), best score (0hard/-3317soft), score calculation speed (1169/sec), phase total (2), environment mode (NON_REPRODUCIBLE).
Is there a sufficient way to restrict steps from using pre-used value ranges? Would there be any draw backs?
I want to use multithread function in optaplaner.
I edited cloudBalancingSolverConfig.xml
and set multithread count 4
but I think it does not work.
All thread name is the same.
19:37:12.980 [l-4-thread-1] DEBUG CH step (1533), time spent (5839), score (-3266init/0hard/-909220soft), selected move count (1276), picked move (CloudProcess-1262 {null -> CloudComputer-1112}).
19:37:12.982 [l-4-thread-1] DEBUG CH step (1534), time spent (5841), score (-3265init/0hard/-909710soft), selected move count (1600), picked move (CloudProcess-1053 {null -> CloudComputer-908}).
19:37:12.984 [l-4-thread-1] DEBUG CH step (1535), time spent (5843), score (-3264init/0hard/-909710soft), selected move count (1277), picked move (CloudProcess-1049 {null -> CloudComputer-1138}).
19:37:12.987 [l-4-thread-1] DEBUG CH step (1536), time spent (5846), score (-3263init/0hard/-912110soft), selected move count (1600), picked move (CloudProcess-1003 {null -> CloudComputer-1231}).
19:37:12.990 [l-4-thread-1] DEBUG CH step (1537), time spent (5849), score (-3262init/0hard/-912110soft), selected move count (1278), picked move (CloudProcess-941 {null -> CloudComputer-1231}).
19:37:12.992 [l-4-thread-1] DEBUG CH step (1538), time spent (5851), score (-3261init/0hard/-912110soft), selected move count (1278), picked move (CloudProcess-893 {null -> CloudComputer-1231}).
19:37:12.994 [l-4-thread-1] DEBUG CH step (1539), time spent (5853), score (-3260init/0hard/-914510soft), selected move count (1600), picked move (CloudProcess-840 {null -> CloudComputer-1286}).
19:37:12.996 [l-4-thread-1] DEBUG CH step (1540), time spent (5855), score (-3259init/0hard/-916910soft), selected move count (1600), picked move (CloudProcess-820 {null -> CloudComputer-1311}).
19:37:12.998 [l-4-thread-1] DEBUG CH step (1541), time spent (5857), score (-3258init/0hard/-916910soft), selected move count (1278), picked move (CloudProcess-779 {null -> CloudComputer-1231}).
19:37:13.000 [l-4-thread-1] DEBUG CH step (1542), time spent (5859), score (-3257init/0hard/-916910soft), selected move count (1279), picked move (CloudProcess-768 {null -> CloudComputer-1286}).
19:37:13.003 [l-4-thread-1] DEBUG CH step (1543), time spent (5862), score (-3256init/0hard/-917470soft), selected move count (1600), picked move (CloudProcess-739 {null -> CloudComputer-750}).
19:37:13.005 [l-4-thread-1] DEBUG CH step (1544), time spent (5864), score (-3255init/0hard/-917470soft), selected move count (1278), picked move (CloudProcess-728 {null -> CloudComputer-1231}).
19:37:13.007 [l-4-thread-1] DEBUG CH step (1545), time spent (5866), score (-3254init/0hard/-917470soft), selected move count (1278), picked move (CloudProcess-671 {null -> CloudComputer-1231}).
19:37:13.009 [l-4-thread-1] DEBUG CH step (1546), time spent (5868), score (-3253init/0hard/-917470soft), selected move count (1279), picked move (CloudProcess-615 {null -> CloudComputer-1286}).
I checked rebased implementation in cloudChangeMove, couldSwapMove.
It was already implemented like belows.
#Override
public CloudComputerChangeMove rebase(ScoreDirector<CloudBalance> destinationScoreDirector) {
return new CloudComputerChangeMove(destinationScoreDirector.lookUpWorkingObject(cloudProcess),
destinationScoreDirector.lookUpWorkingObject(toCloudComputer));
}
Did I forget something?
The version of optaplaner is 7.23.
But it was the same previous version(7.22)
Please let me know my problem.
Turn on TRACE logging. The steps are done by the solver thread, the move evaluations (only visible in trace logging) are done by the move threads.
I have set up a solver with two local search phases and it works fine. However, there was a time that the 2nd phase didn't make any move in about 1 minute, as the log shows below:
...
2016-05-07 14:14:55,847 [main] DEBUG LS step (10069), time spent (593822), score (0hard/-81medium/5395020soft), best score (0hard/-80medium/5393781soft), accepted/selected move count (5/48), picked move (CL [cID=1147576, id=27246 => SL [id=49, sID=E942648]] <=> CL [cID=1133912, id=14716 => SL [ id=7, sID=E942592]]).
2016-05-07 14:14:55,858 [main] DEBUG LS step (10070), time spent (593833), score (0hard/-81medium/5395390soft), best score (0hard/-80medium/5393781soft), accepted/selected move count (5/18), picked move (CL [cID=1142322, id=22533 => SL [ id=51, sID=E943251]] <=> CL [cID=1134362, id=14118 => SL [ id=49, sID=E942648]]).
2016-05-07 14:14:55,858 [main] INFO Local Search phase (1) ended: step total (10071), time spent (593833), best score (0hard/-80medium/5393781soft).
2016-05-07 14:16:05,042 [main] INFO Local Search phase (2) ended: step total (0), time spent (663017), best score (0hard/-80medium/5393781soft).
2016-05-07 14:16:05,042 [main] INFO Solving ended: time spent (663017), best score (0hard/-80medium/5393781soft), average calculate count per second (2771).
Before phase 1 ended, there wasn't any improvement in the last couple of steps. And phase 2 started but made 0 step in a minute. The solver then ended because it has reached the maximum time allowed.
I'm a bit surprised that phase 2 made no step at all. Is it simply because it didn't manage to find any better score?
If you don't see any moves in TRACE log (as suggested in comments), it could be because you're using a custom move list factory and that's taken too long to initialize.
I was looking at the examples found on this website :
http://www.tutorialspoint.com/operating_system/os_process_scheduling_algorithms.htm
And there's something that just doesn't make sense about those examples. Take shortest-job-first for example. The premise is that you take the process with the least execution time and run that first.
The example runs p1 first and then p0. But WHY? At t = 0 the only process that exists in the queue is p0. Wouldn't that start running at t = 0, and then p1 would start at t = 6?
I've got the same issue with priority based scheduling.
you are right , since the process P0 has arrived at the queue at 0 sec and before P1 , it will start executing before P1 .
Their answer would be correct if there was no arrival time for the corresponding process and in that case , it is considered that all the processes have reached at the queue at the same time .So, the process with shortest executing time will be executed by CPU first .
I ran into a situation where only two valid moves that is allowed happens during the construction phase. This results in a situation where there is no valid moves to perform during local search phase. All the moves tried during the local search phase is not doable.
This results in NullPointerException at the end of local search phase.
Exception in thread "main" java.lang.NullPointerException
at org.optaplanner.core.impl.solver.scope.DefaultSolverScope.getBestScoreWithUninitializedPrefix(DefaultSolverScope.java:205)
at org.optaplanner.core.impl.solver.DefaultSolver.outerSolvingEnded(DefaultSolver.java:216)
at org.optaplanner.core.impl.solver.DefaultSolver.solve(DefaultSolver.java:161)
at com.example.jobplanner.Application.main(Application.java:328)
Here is the trace of the two valid moves that happens during the construction phase.
CH step (0), time spent (151), score (0hard/0soft), selected move count (1), picked move (com.example.jobplanner.domain.ServiceJobChip#981500 => com.example.jobplanner.domain.ShopFloorBay#9da471).
Update [fact 0:5:4216052:4216052:10:DEFAULT:NON_TRAIT:com.example.jobplanner.domain.ServiceJobChip#4054f4]
CH step (1), time spent (158), score (0hard/0soft), selected move count (1), picked move (com.example.jobplanner.domain.ServiceJobChip#4054f4 => com.example.jobplanner.domain.ShopFloorBay#15ac5d5).
Construction Heuristic phase (0) ended: step total (2), time spent (159), best score (null).
The model of the domain I am trying to optimize would be something like:
Each bay has --> set of tasks that can be performed
Job -> collection of tasks to be performed
Task --> can be scheduled to bays which can perform the task