monad-dataplane: share pacing between udp and tcp - #3237
Draft
dshulyak wants to merge 1 commit into
Draft
Conversation
dshulyak
force-pushed
the
dmitry/tcp-pacing
branch
2 times, most recently
from
September 1, 2026 12:50
1118450 to
eac320a
Compare
dshulyak
force-pushed
the
dmitry/tcp-pacing
branch
from
September 2, 2026 06:07
eac320a to
aa3c40b
Compare
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
the distributed key generation protocol must send about 1 gib within 30 minutes while consensus traffic shares the same network. its messages can reach 2 mib, so retransmitting complete messages after packet loss would be wasteful. the required average dkg throughput is approximately 4.77 mbps.
dkg traffic uses background tcp through the shared global pacer, while consensus udp traffic receives high priority. tcp consumes otherwise-idle bandwidth and yields during consensus bursts. tcp also retransmits only missing network segments, avoiding full-message retransmission during packet loss.
the test used a 1 gbps pacing limit with ten tcp receivers and a 50 mib udp burst every two seconds. metrics were polled externally every 10 ms. tcp averaged 760 mbps over the run and 969 mbps while udp was idle, with zero drops or errors. this provides substantially more throughput than the dkg protocol requires.