After browsing a bit through the UVM source code, I noticed that singletons for each phase exist. For the run phase this can be accessed by calling uvm_run_phase::get().
I tried calling uvm_run_phase::get() from both the run_phase and the end_of_elaboration phase and as expected both calls returned a handle to the same object. I am a bit puzzled as the object returned by this method is not the same object that is passed to the run_phase(...) task as it's phase argument.
I've created an example on EDA Playground: http://www.edaplayground.com/x/2PL
Does anyone know exactly what the phase argument getting passed to phase methods is and why it's not the singleton?
After asking the question on the UVM forum, I've found out that there are multiple instances of uvm_phase for each phase that get created, one being used for scheduling and the other being used as the actual implementation.
Calling uvm_run_phase::get() will return the implementation node for the run phase, whereas the argument being past to run_phase(...) is the scheduling node. To get the scheduling node from any other phase, one can use phase.find_by_name("run",0).
Here is the link to the original thread in the UVM forum: http://forums.accellera.org/topic/1769-uvm-phase-singletons/
Related
I've stumbled upon a problem when using some more complex argo workflows with initialization and clean-up logic. We are running some initialization in one of the initial steps of the workflow (e.g. creation of some resources) and we'd like to perform a clean-up regardless of the status of the workflow. onExit template seems to be an ideal solution (I think that clean-up is even mentioned in argo documentation as predestined for tasks of the onExit template).
However, I haven't found a way yet to pass some values to it. For example - let's say that in the initialization phase we created some resource with id some-random-unique-id and we'd like to let the onExit container know what resources it needs to clean up.
We tried the outputs of some steps, but it seems that steps are unknown in the onExit template.
Is there a built-in argo mechanism to pass this kind of data? We'd like to avoid some external services (like key-value storage service that would hold the context).
You can mark output parameters as global using the globalName field. A global output parameter, assuming it has been set, can be accessed from anywhere in the Workflow, including in an exit handler.
The example file for writing and consuming global output parameters should contain all the information you need to use global output parameters in an exit handler.
https://github.com/argoproj/argo-workflows/blob/master/examples/global-outputs.yaml
Please share if you found a cleaner solution.
I see this workflow example for exit handler with parameters, which may resolve your issue: Exit Handler with Parameters
One method I found was to add steps to the template I want to call the exit template on, and call the exit template as the last step so I can access steps variables. I changed the original template to an inline template so the structure does not change too much, just involves adding the last step call to the exit template. Unfortunately, this does not utilize the actual onExit method.
Inline steps template example
1.https://sourcegraph.com/github.com/argoproj/argo-workflows/-/blob/examples/exit-handler-with-param.yaml?L21
2.https://raw.githubusercontent.com/argoproj/argo-workflows/master/examples/steps-inline-workflow.yaml
Recently, I am studying the problem of task scheduling in Flink. My purpose is to schedule subtasks to a slot of the specified node according to my own needs by modifying some source codes of the scheduling part. Through remote debugging and checking the source code, I found the following method call stack, most of which I can't understand (the comments are a little less), especially in this method: org.apache.flink.runtime.jobmaster.slotpool.SchedulerImpl#allocateMultiTaskSlot. I guess the code that allocates slots to tasks is around here. Because it is too difficult to read the source code, I have to ask you for help. Of course, if there is a better way to achieve my needs, please specify one or two. Sincerely look forward to your reply! Thank you very much!!!
The method call stack is as follows(The version of Flink I use is 1.11.1):
org.apache.flink.runtime.jobmaster.JobMaster#startJobExecution
org.apache.flink.runtime.jobmaster.JobMaster#resetAndStartScheduler
org.apache.flink.runtime.jobmaster.JobMaster#startScheduling
org.apache.flink.runtime.scheduler.SchedulerBase#startScheduling
org.apache.flink.runtime.scheduler.DefaultScheduler#startSchedulingInternal
org.apache.flink.runtime.scheduler.strategy.EagerSchedulingStrategy#startScheduling
(This is like the method call chain of PipelinedRegionSchedulingStrategy class. In order to simply write it as the method call chain of EagerSchedulingStrategy class, it should have no effect)
org.apache.flink.runtime.scheduler.strategy.EagerSchedulingStrategy#allocateSlotsAndDeploy
org.apache.flink.runtime.scheduler.DefaultScheduler#allocateSlotsAndDeploy
org.apache.flink.runtime.scheduler.DefaultScheduler#allocateSlots
org.apache.flink.runtime.scheduler.DefaultExecutionSlotAllocator#allocateSlotsFor
org.apache.flink.runtime.executiongraph.SlotProviderStrategy.NormalSlotProviderStrategy#allocateSlot
org.apache.flink.runtime.jobmaster.slotpool.SchedulerImpl#allocateSlot
org.apache.flink.runtime.jobmaster.slotpool.SchedulerImpl#allocateSlotInternal
org.apache.flink.runtime.jobmaster.slotpool.SchedulerImpl#internalAllocateSlot
org.apache.flink.runtime.jobmaster.slotpool.SchedulerImpl#allocateSharedSlot
org.apache.flink.runtime.jobmaster.slotpool.SchedulerImpl#allocateMultiTaskSlot
(I feel that this is the key to allocate slot for subtask, that is, execution vertex, but there is no comment, and I don't understand the process idea, so I can't understand it.)
For the use case of being able to abort parallel simulations with a MATLAB GUI, I would like to stop all scheduled simulations after the user pressed the Stop button.
All simulations are submitted at once using the parsim command, hence something like a callback to my GUI variables (App Designer) would be the most preferable solution.
Approaches I have considered but were not exactly providing a desirable solution:
The Simulation Manager provides the functionality to close simulations using its own interface. If I only had the code its Stop button executes...
parsim uses the Simulink.SimulationInput class as input to run simulations, allowing to modify the preSimFcn at the beginning of each simulation. I have not found a way to "skip" the simulation at its initialization phase apart from intentionally throwing an error so far.
Thank you for your help!
Update 1: Using the preSimFcn to set the the termination time equal to the start time drastically reduces simulation time. But since the first step still is computed there has to be a better solution.
simin = simin.setModelParameter('StopTime',get_param(mdl,'StartTime'))
Update 2: Intentionally throwing an error executing the preSimFcn, for example by setting it to
simin = simin.setModelParameter('SimulationCommand','stop')
provides the shortest termination times for me so far. Though, it requires catching and identifying the error in the ErrorMessageof the Simulink.SimulationOutput object. As this is exactly the "ugly" implementation I wanted to avoid, the issue is still active.
If you are using 17b or later, parsim provides an option to 'RunInBackground'. It returns an array of Future objects.
F = parsim(in, 'RunInBackground', 'on')
Please note that is only available for parallel simulations. The Simulink.Simulation.Future object F provides a cancel method which will terminate the simulation. You can use the fetchOutputs methods to fetch the output from the simulation.
F.cancel();
I wish to measure the time it takes for some of my Actor methods to execute. I plan to use OnPreActorMethodAsync and OnPostActorMethodAsync to log when these methods have begun and finished execution.
However I also wish to measure the time it takes for these methods to execute. My original plan was to use an instance of StopWatch to begin timing in the OnPreActorMethodAsync class and stop in the OnPostActorMethodAsync class. However I'm unsure how I can accessthe same StopWatch instance in the second method. If there is a better way of doing this I would be very interested.
If anyone could point me in the right direction on this matter that would be great.
Many Thanks
By using some dirty reflection tricks you can access the Actors' call context.
More about it here: https://stackoverflow.com/a/39526751/5946937
(Disclaimer: Actor internals may have changed since then)
We are developing an ABM under AnyLogic 7 and are at the point where we want to make multiple simulations from a single experiment. Different parameters are to be set for each simulation run so as to generate results for a small suite of standard scenarios.
We have an experiment that auto-starts without the need to press the "Run". Subsequent pressing of the Run does increment the experiment counter and reruns the model.
What we'd like is a way to have the auto-run, or single press of Run, launch a loop of simulations. Within that loop would be the programmatic adjustment of the variables linked to passed parameters.
EDIT- One wrinkle is that some parameters are strings. The Optimization or Parameter Variation experiments don't lend themselves to enumerating a set of strings to be be used across a set of simulation runs. You can set a string per parameter for all the simulation runs within one experiment.
We've used the help sample for "Running a Model from Outside Without Presentation Window", to add the auto-run capability to the initial experiment setup block of code. A method to wait for Run 0 to complete, then dispatch Run 1, 2, etc, is needed.
Pointers to tutorial models with such features, or to a snip of code for the experiment's java blocks are much appreciated.
maybe I don't understand your need but this certainly sounds like you'd want to use a "Parameter Variation" experiment. You can specify which parameters should be varied in which steps and running the experiment automatically starts as many simulation runs as needed, all without animation.
hope that helps
As you, I was confronted to this problem. My aim was to use parameter variation with a model and variation were on non numeric data, and I knew the number of runs to start.
Then i succeed in this task with the help of Custom Variation.
Firstly I build an experiment typed as 'multiple run', create my GUI (user was able to select the string values used in each run.
Then, I create a new java class which inherit from the previous 'multiple run' experiment,
In this class (called MyMultipleRunClass) was present:
- overload of the getMaximumIterations method from default experiment to provide to default anylogic callback the correct number of iteration, and idnex was also used to retrieve my parameter value from array,
- implementation of the static method start,
public static void start() {
prepareBeforeExperimentStart_xjal( MyMultipleRunClass.class);
MyMultipleRunClass ex = new MyMultipleRunClass();
ex.setCommandLuneArguments_xjal(null);
ex.setup(null);
}
Then the experiment to run is the 'empty' customExperiment, which automatically start the other Multiple run experiment thru the presented subclass.
Maybe it exists shortest path, but from my point of view anylogic is correctly used (no trick with non exposed interface) and it works as expected.