Initial algorithm for nominating valid 'parallel' tx sets. #4486
Conversation
This file contains 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
dmkozh
force-pushed
the
parallel_txset_nomination
branch
6 times, most recently
from
f911313 to
46e2e52
Compare
6 tasks
dmkozh
force-pushed
the
parallel_txset_nomination
branch
from
46e2e52 to
ba6137b
Compare
dmkozh
changed the title
This is a relatively simple algorithm, but it should be serviceable for traffic with relatively low amount of transitive IO conflicts.
dmkozh
force-pushed
the
parallel_txset_nomination
branch
from
ba6137b to
8afa754
Compare
sisuresh
reviewed
Feb 25, 2025
| { | ||
| ZoneScoped; | ||
| // A fast-fail condition to ensure that adding the transaction won't | ||
| // exceed the theorethical limit of instructions per stage. |
There was a problem hiding this comment.
Suggested change
| // exceed the theorethical limit of instructions per stage. | |
| // exceed the theoretical limit of instructions per stage. |
| std::vector<std::shared_ptr<Cluster const>> | ||
| createNewClusters(BuilderTx const& tx, | ||
| std::unordered_set<Cluster const*> const& txConflicts, | ||
| bool& packed) |
There was a problem hiding this comment.
Instead of returning packed as a ref parameter, why not return a pair eg. std::pair<std::vector<std::shared_ptr<Cluster const>>, bool>?
There was a problem hiding this comment.
That works as well, I just wanted to avoid copying the vector unnecessarily.
| // This is not configurable for now. It doesn't need to be a network-wide | ||
| // setting, but on the other hand there aren't many good values for it and | ||
| // it's not clear what the right way to configure it would be, if at all. | ||
| SOROBAN_PHASE_STAGE_COUNT = 1; |
There was a problem hiding this comment.
I think you've mentioned this setting should be dynamically determined by the contents of the tx queue instead of hardcoded. What's the plan wrt that?
There was a problem hiding this comment.
I think that would be a separate PR. But I think we could make a few tx sets in parallel and use the one with the smallest amount of stages that provides roughly the same utilization.
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
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.
Initial algorithm for nominating valid 'parallel' tx sets.
This is a relatively simple algorithm, but it should be serviceable for traffic with relatively low amount of transitive IO conflicts.