Influence of runspeed

L.S.

In my model the strange thing happens that the runspeed of my model influences the performance. In short, when I run it at a time below say 500x real time then everything goes as expected. However when I speed up a little more the model is not behaving like expected anymore and my production cannot be completed because at some point there seem to be no messages anymore which keep the model running...
Is anybody familiar with this problem or can point out some possible mistake of mine? Preferably I do not share my model with the whole community.

The only issue I had once was that I checked the xloc() of an item on a conveyor. If you do stuff like this you should be familiar with the command updatelocations().

But from your post it sound, that you use messages for control. Maybe the problem is how and where (which trigger) you send the messages. Without more information it is difficult to think about the possibilites.

I hope that I can expect that you know the difference between sendmessage() and senddelayedmessage() in zero time units.

Anyway, that's all what comes directly to my mind.

I'm curious. Do you pass a flowitem pointer like tonum(item) in your messages? If so, how is it used in the OnMessage?

Hi,

After a lot of searching I found a mistake in a tasksequence which seems to prevent the problem I mentioned because now I can run at unlimited speed. However, if I leave out this statement (an extra message as to prevent the model from running out of messages) the problem exists and I find that strange. Thus, maybe I am still making a fundamental (stupid) mistake at which I should be glad to be pointed at. I added the tasksequence which is in my opinion the source of the troubles. The line with all the asterisks I added lately and seems to help.

Maurits,

Besides the update locations of Tom, we have also find this problem when using kinematics. I don't know if you use that? If so, check if you use updatekinematics enough.


Just the code is for me at least not enough to figure it out. Otherwise if you have it again, send us the model and we will see if we can find it.

Steven Hamoen
Talumis

Hi,
I have solved the problem now completely, and I think I found the source of the problem. At some point in the model I used the value from a non existing column in a global table. I suppose this caused a randomness which sometimes resulted in the problem. Despite my own faults I still found it a strange problem :-)

Hi Maurits,


Good to hear that you found it.

Yes these things are very strange, but what internally in the memory of the computer is happening is that you somewhere change a memory position that can be used by anything else in the program so that it is not clear what exactly is influenced by your actions. That it is why it is so strange and unpredictable.

Steven