Closed catalinii closed 1 year ago
Hi @catalinii ,
This patch seems interesting. You have the idea to update the satipc
input module with this new FIFO implementation? And what about the processing internal buffer?
I was thinking about using fifo for the satipc module (for sending the messages instead of the hash table), but I think a better option is to refactor and prevent the satipc module from buffering responses. This would be better because of a message is buffered any update to the list of pids for example, requires a new message. If there is no buffering the satipc module can just share it's current state rather than an older state. I plan next to replace the socket buffering with FIFO, as based on current consumption it uses 2x-4x more memory that usable.