The rest traffic has a delay of ~250ms per message per node and then it rests the full delay time (TRAFFIC_DELAY_SECONDS).
The ~250ms delay is due to the time it takes to generate the proof.
If there are many nodes or a very short delay is configured then the message rate can be far off from the expected.
This can cause confusion and make it seem as if there is message loss.
Possible ways to improve the delay mismatch:
- send messages async
- implement sending messages in batches to limit the number of messages sent at once)
- decrease the sleep time by the time taken to send the message
The rest traffic has a delay of ~250ms per message per node and then it rests the full delay time (
TRAFFIC_DELAY_SECONDS).The ~250ms delay is due to the time it takes to generate the proof.
If there are many nodes or a very short delay is configured then the message rate can be far off from the expected.
This can cause confusion and make it seem as if there is message loss.
Possible ways to improve the delay mismatch: