LabVIEW
While Loops are not working in parallel
Solved!
HI Abdelrahman.
Sounds a bit like you've got some non-reentrant functions within your while loops. where only one instance of the code is allowed to be executed at any time. Have a look at this: Reentrancy: Allowing Simultaneous Calls to the Same SubVI - NI
If that there is probably some other recourse that the two loops are fighting over.
Beyond that, I'm not able to assist without seeing some code. (I don't have access to anything newer than 2020 atm)
If I'm not mistaken:
Shared clone, will spawn more clones as needed.
Preallocated will look at your code and spawn one per call.
I normally use shared clone, unless I'm on a Real-time system where the occasional Jitter of spawning a clone is not acceptable.
@LabVIEWFairy wrote:
If I'm not mistaken:
Shared clone, will spawn more clones as needed.
Preallocated will look at your code and spawn one per call.
I normally use shared clone, unless I'm on a Real-time system where the occasional Jitter of spawning a clone is not acceptable.
Shared Clone will create a new clone as is needed. It can use less memory as a single clone can be used by multiple instances of the reentrant VI (ex 5 clones could be used for 10 calls if up to 5 are needed at the same time).
Preallocated Clone creates a new clone for every call. This is required if the reentrant needs to maintain state. It also reduces jitter as the clone pool done not need to be querried and possibly use extra time to spin up another clone.
@crossrulz wrote:
@LabVIEWFairy wrote:
If I'm not mistaken:
Shared clone, will spawn more clones as needed.
Preallocated will look at your code and spawn one per call.
I normally use shared clone, unless I'm on a Real-time system where the occasional Jitter of spawning a clone is not acceptable.Shared Clone will create a new clone as is needed. It can use less memory as a single clone can be used by multiple instances of the reentrant VI (ex 5 clones could be used for 10 calls if up to 5 are needed at the same time).
Preallocated Clone creates a new clone for every call. This is required if the reentrant needs to maintain state. It also reduces jitter as the clone pool done not need to be querried and possibly use extra time to spin up another clone.
So when would you not want to maintain state within your clone? I'm curious because I want to expand my knowledge in this area. I've only ever wanted them to act like completely independent VIs. Thanks!
@billko wrote:
So when would you not want to maintain state within your clone? I'm curious because I want to expand my knowledge in this area. I've only ever wanted them to act like completely independent VIs. Thanks!
There needs to be some distinctions here. There are reentrant VIs that are called at the very top level VI to just be a separate parallel thread. These should maintain their own state. But then there are the reentrant VIs that are "constantly" being called in a loop. State should not be maintained in those VIs, but maintained by the loop.
But that's just my opinion.
@crossrulz wrote:
@billko wrote:
So when would you not want to maintain state within your clone? I'm curious because I want to expand my knowledge in this area. I've only ever wanted them to act like completely independent VIs. Thanks!There needs to be some distinctions here. There are reentrant VIs that are called at the very top level VI to just be a separate parallel thread. These should maintain their own state. But then there are the reentrant VIs that are "constantly" being called in a loop. State should not be maintained in those VIs, but maintained by the loop.
But that's just my opinion.
This kind of goes along with the post above yours. I have to consider your responses closely before I say I understand it. I know you think this is trivial, and it probably is. It's just new to me, and I appreciate your patience.
