Upstream entry reconstruction fixes #61
Open
pmantica11 wants to merge 1 commit intojito-labs:masterfrom
Open
Conversation
esemeniuc
reviewed
Dec 3, 2025
| ); | ||
|
|
||
| if should_reconstruct_shreds { | ||
| let _ = reconstruct_tx.try_send(packet_batch.clone()); |
Collaborator
There was a problem hiding this comment.
can we switch this back to try_send, don't to have OOM due to unbounded linked list. lets increase the buffer size instead
esemeniuc
reviewed
Dec 3, 2025
| ) -> usize { | ||
| deshredded_entries.clear(); | ||
| slot_fec_indexes_to_iterate.clear(); | ||
| let mut data_fec_indexes_to_reconstruct = Vec::new(); |
Collaborator
There was a problem hiding this comment.
lets re-use the vec and clear it instead of allocating a new one each time
esemeniuc
reviewed
Dec 3, 2025
| // failed to reconstruct the entries from the index due to never finding the start data boundary. | ||
| // To deal with this possible out-of-order scenario, we retry rebuilding the entrie(s) for the next | ||
| // fec index whenever we encounter a data complete boundary. | ||
| if let Some(next_slot) = &state_tracker.data_shreds[index + 1] { |
Collaborator
There was a problem hiding this comment.
im a bit confused by this logic, state_tracker.data_shreds only tracks shreds for a single slot. how does this give you next slot information?
Collaborator
There was a problem hiding this comment.
also we only ever push to this when shreds come out of order. if they always come in order, the never push to the vec, and line 186 for (slot, fec_set_index) in data_fec_indexes_to_reconstruct.iter() { wouldnt iterate on anything
Collaborator
|
Please run the test suite, the new changes decodes less transactions than before |
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
2 participants
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.
Overview
In this PR, I replace multiple issues causing deshredding to fail:
try_sendwithsendto ensure that shreds are delivered to the reconstructor thread. (Without this, I got a lot of missing shreds in the reconstructor thread).Tests
Verified locally that we didn't get any missing transactions for over 1000 slots after the two fixes above.