Open baumanj opened 5 years ago
The worst-case message round-trip time (how long the algorithm waits for ack
messages) should probably also be made tunable as part of this work.
An environment variable-based approach will be trivial with our EnvConfig abstraction.
Dynamic self-tuning is left as an exercise for the reader 😄
This issue has been automatically marked as stale because it has not had recent activity. It will be closed if no further activity occurs. We value your input and contribution. Please leave a comment if this issue still affects you.
This issue has been automatically marked as stale because it has not had recent activity. It will be closed if no further activity occurs. We value your input and contribution. Please leave a comment if this issue still affects you.
Since the relationship between protocol period Tʹ and connection latencies is vital to achieving the properties described in the SWIM paper, a facility for tuning the protocol period after compilation would be useful. #5975 describes how we should make this tunable at runtime.
If we determine that dynamic changes during runtime are necessary, we should implement a self-tuning value. The runtime could determine what the threshold should be based on the size of the mesh network and the latencies it's experiencing.