Stop atoms when running

[Tommytom]

Hi there,

I have a small question concerning the ED model I've created. Is it possible to stop the specific operation of an atom during the simulation.

In my model, a server is being active. However, I want to stop the operation immediately when another event occurs. I want to stop the server from being active and sent the content to the queue behind the server. After another operation, I want to activate the server again.
This problem can also be translated to the MTBF MTTR availability atom. In a realistic situation, I can image that the server is being maintained before its failure, but due to another event (for example: failure in another server) its maintenance is rescheduled. But in the simulation environment, the MTBF MTTR availability input cannot handle it, I believe. This problem would be solved when it is possible to stop the process of the atom, sent the content to the output, and start the atom again when the user demands it.

Whenever I'm wrong, and this is possible instead, I would like to hear it!

Tom

[Josie]

Dear Tom,
Could you tell me more detail about the situation or environment you want to build the model? If you just need to deactivate the server and product can be unfinished, just move the products in this server away (Using MoveAtom), and destroy all the events that have been schedule in it.
If you need to continue dealing with the unfinished products after this server can work again, one solution is just to keep products in the deactivated server, don't let them go and processing them again once the server is activated. Another solution, move the unfinished products to other places, such as a queue like you mentioned, and when the server is fine, move products back, continuing processing products in that server.
In later solution, a example model has been composed by me, it goes through the following steps:
1. close the input channels of it
2. move the products in it to the later queue
3. destroy all the events that have been scheduled in this server Atom.
4. when it is time and this server can be activated again, move the products back from the queue which stores the moved products, and then let the server continue what it should do.

Please find attachment the example model. The most important atoms are the two servers: "Server3" and "Server9".
Server3 is the server which you want to deactivate and activate from time to time, and Server9 functions as a controlling block. Important controlling codes are written in atoms.

Best regards,
Josie

[Tommytom]

Hi Josie,

Thanks for your solution! This works fine in my simulation too