fix: prevent deadlock in concurrent BERT auto-embeddings - #206
Merged
Merged
Conversation
Concurrent requests can enter the shared Rayon pool and block on the BERT model mutex while the active forward pass needs that pool to finish. Nested Rayon execution can therefore deadlock inference. Serialize conversions per shared BERT model before entering the pool, while retaining parallel computation within each conversion. Add FFI regression coverage for concurrent 500-document MiniLM batches. Related issue: manticoresoftware/manticoresearch#4915
- Move model locking outside Rayon prediction tasks - Pass locked state through BERT, T5, and quantized paths - Remove per-forward mutex locking from embedding predictions
donhardman
approved these changes
Sep 18, 2026
donhardman
left a comment
donhardman
left a comment
Member
There was a problem hiding this comment.
The fix is a workaround, not a fix. So I implemented proper changes that should eliminate it but still subject of check on original problem. Now its clean fix that applied to all models
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Concurrent requests can enter the shared Rayon pool and block on the BERT model mutex while the active forward pass needs that pool to finish. Nested Rayon execution can therefore deadlock inference.
Serialize conversions per shared BERT model before entering the pool, while retaining parallel computation within each conversion. Add FFI regression coverage for concurrent 500-document MiniLM batches.
Related issue: manticoresoftware/manticoresearch#4915
Related PR: manticoresoftware/manticoresearch#4916