zhannngchen commented on code in PR #12605:
URL: https://github.com/apache/doris/pull/12605#discussion_r971671230
##########
be/src/olap/memtable_flush_executor.h:
##########
@@ -107,7 +107,7 @@ class MemTableFlushExecutor {
void init(const std::vector<DataDir*>& data_dirs);
Status create_flush_token(std::unique_ptr<FlushToken>* flush_token,
RowsetTypePB rowset_type,
- bool is_high_priority);
+ bool merge_on_write, bool is_high_priority);
Review Comment:
the new parameter should rename to `serial_mode`, MoW is a feature of unique
key table, use `merge_on_write` here is confusing.
##########
be/src/olap/memtable_flush_executor.cpp:
##########
@@ -132,10 +132,12 @@ void MemTableFlushExecutor::init(const
std::vector<DataDir*>& data_dirs) {
}
// NOTE: we use SERIAL mode here to ensure all mem-tables from one tablet are
flushed in order.
+// unique key merge on write should flush serial cause calc delete bitmap
should load segment serial
Status MemTableFlushExecutor::create_flush_token(std::unique_ptr<FlushToken>*
flush_token,
- RowsetTypePB rowset_type,
bool is_high_priority) {
+ RowsetTypePB rowset_type,
bool merge_on_write,
+ bool is_high_priority) {
if (!is_high_priority) {
- if (rowset_type == BETA_ROWSET) {
+ if (rowset_type == BETA_ROWSET && !merge_on_write) {
Review Comment:
L150 should also consider the parameter.
--
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]
