Fix broken parallelism in quic-client (#2526)
* Fix broken parallelism in quic-client
Fixes excessive fragmentation by TPU clients leading to a large
number of streams per conn in 'sending' state simultaneously.
This, in turn, requires excessive in-memory buffering server-side
to reassemble fragmented transactions.
- Simplifies QuicClient::send_batch to enqueue send operations
in sequential order
- Removes the "max_parallel_streams" config option
The quic-client now produces an ordered fragment stream when
scheduling send operations from a single-thread.
* quic-client: remove outdated test
---------
Co-authored-by: Richard Patel <ripatel@jumptrading.com>
Co-authored-by: Alessandro Decina <alessandro.d@gmail.com>
(cherry picked from commit
|
||
---|---|---|
.. | ||
src | ||
tests | ||
Cargo.toml |