kpfly commented on code in PR #11939:
URL: https://github.com/apache/doris/pull/11939#discussion_r956662011
##########
be/src/exec/tablet_sink.cpp:
##########
@@ -302,8 +302,7 @@ Status NodeChannel::add_row(Tuple* input_tuple, int64_t
tablet_id) {
// _cancelled may be set by rpc callback, and it's possible that
_cancelled might be set in any of the steps below.
// It's fine to do a fake add_row() and return OK, because we will check
_cancelled in next add_row() or mark_close().
while (!_cancelled && _pending_batches_num > 0 &&
- (_pending_batches_bytes > _max_pending_batches_bytes ||
-
_parent->_mem_tracker->limit_exceeded(_max_pending_batches_bytes))) {
Review Comment:
> this is not only useless but a performance bug which will cause
unnecessary sleep.
Is there any performance tests?
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: [email protected]
For queries about this service, please contact Infrastructure at:
[email protected]
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
