github-actions[bot] commented on code in PR #65362:
URL: https://github.com/apache/doris/pull/65362#discussion_r3548417788
##########
be/src/load/group_commit/wal/wal_table.cpp:
##########
@@ -122,6 +128,9 @@ Status WalTable::_relay_wal_one_by_one() {
for (auto retry_wal_info : need_retry_wals) {
_replay_wal_map.emplace(retry_wal_info->get_wal_path(),
retry_wal_info);
}
+ if (_replay_wal_map.empty()) {
Review Comment:
The failure reason is cleared only when the whole table replay map becomes
empty. That leaves stale state behind if a WAL that previously failed later
replays successfully while other WALs are still pending or not yet due for
retry; `_check_wal_backlog()` then sees `wal_num >= limit` plus this non-empty
old reason and rejects new async loads even though replay is no longer failing.
This breaks the new intended distinction between temporary WAL accumulation and
a failure-blocked backlog. Please clear the reason when the WAL that produced
it succeeds, or track failure state per WAL so the admission check only
considers currently failing WALs.
--
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]
