This is an automated email from the ASF dual-hosted git repository.
sollhui pushed a commit to branch master
in repository https://gitbox.apache.org/repos/asf/doris.git
The following commit(s) were added to refs/heads/master by this push:
new 43cf4fd43af [fix](auto partition) keep load row metrics monotonic for
auto partition (#64109)
43cf4fd43af is described below
commit 43cf4fd43af1a836c7fedef5aa462e86fa93a2c7
Author: hui lai <[email protected]>
AuthorDate: Fri Jun 5 14:14:16 2026 +0800
[fix](auto partition) keep load row metrics monotonic for auto partition
(#64109)
### What problem does this PR solve?
When loading into an auto partition table, BE may temporarily batch rows
whose target partitions do not exist yet. The previous logic first
incremented `doris_be_load_rows` when receiving the input block, then
decremented it when those rows were moved into the auto-partition
batching block, and finally incremented it again when the batched rows
were replayed after partition creation.
Because `doris_be_load_rows` is exposed as a counter, the negative
adjustment made the metric non-monotonic and could show row count drops
during auto partition loads.
This PR keeps the existing `RuntimeState` row compensation logic
unchanged, but stops applying negative deltas to the global BE load
row/byte metrics. It also skips global metric increments when replaying
the internal batched block, so each source row is counted once.
---
be/src/exec/sink/vrow_distribution.cpp | 3 ---
be/src/exec/sink/writer/vtablet_writer.cpp | 7 +++++--
be/src/exec/sink/writer/vtablet_writer_v2.cpp | 7 +++++--
3 files changed, 10 insertions(+), 7 deletions(-)
diff --git a/be/src/exec/sink/vrow_distribution.cpp
b/be/src/exec/sink/vrow_distribution.cpp
index 891c4b394cb..96cf8966fe0 100644
--- a/be/src/exec/sink/vrow_distribution.cpp
+++ b/be/src/exec/sink/vrow_distribution.cpp
@@ -27,7 +27,6 @@
#include "common/cast_set.h"
#include "common/logging.h"
-#include "common/metrics/doris_metrics.h"
#include "common/status.h"
#include "core/assert_cast.h"
#include "core/column/column.h"
@@ -416,8 +415,6 @@ Status VRowDistribution::_deal_missing_map(const Block&
input_block, Block* bloc
rows_stat_val -= new_bt_rows - old_bt_rows;
_state->update_num_rows_load_total(old_bt_rows - new_bt_rows);
_state->update_num_bytes_load_total(old_bt_bytes - new_bt_bytes);
- DorisMetrics::instance()->load_rows->increment(old_bt_rows - new_bt_rows);
- DorisMetrics::instance()->load_bytes->increment(old_bt_bytes -
new_bt_bytes);
return Status::OK();
}
diff --git a/be/src/exec/sink/writer/vtablet_writer.cpp
b/be/src/exec/sink/writer/vtablet_writer.cpp
index 0eb98ae48dd..47b67cbfc9e 100644
--- a/be/src/exec/sink/writer/vtablet_writer.cpp
+++ b/be/src/exec/sink/writer/vtablet_writer.cpp
@@ -2064,6 +2064,7 @@ Status VTabletWriter::write(RuntimeState* state,
doris::Block& input_block) {
// check out of limit
RETURN_IF_ERROR(_send_new_partition_batch());
+ const bool is_replaying_batched_block = _row_distribution._deal_batched;
auto rows = input_block.rows();
auto bytes = input_block.bytes();
if (UNLIKELY(rows == 0)) {
@@ -2078,8 +2079,10 @@ Status VTabletWriter::write(RuntimeState* state,
doris::Block& input_block) {
// the real 'num_rows_load_total' will be set when sink being closed.
_state->update_num_rows_load_total(rows);
_state->update_num_bytes_load_total(bytes);
- DorisMetrics::instance()->load_rows->increment(rows);
- DorisMetrics::instance()->load_bytes->increment(bytes);
+ if (!is_replaying_batched_block) {
+ DorisMetrics::instance()->load_rows->increment(rows);
+ DorisMetrics::instance()->load_bytes->increment(bytes);
+ }
_row_distribution_watch.start();
RETURN_IF_ERROR(_row_distribution.generate_rows_distribution(
diff --git a/be/src/exec/sink/writer/vtablet_writer_v2.cpp
b/be/src/exec/sink/writer/vtablet_writer_v2.cpp
index 31a2d788192..06e35210635 100644
--- a/be/src/exec/sink/writer/vtablet_writer_v2.cpp
+++ b/be/src/exec/sink/writer/vtablet_writer_v2.cpp
@@ -493,6 +493,7 @@ Status VTabletWriterV2::write(RuntimeState* state, Block&
input_block) {
// check out of limit
RETURN_IF_ERROR(_send_new_partition_batch());
+ const bool is_replaying_batched_block = _row_distribution._deal_batched;
auto input_rows = input_block.rows();
auto input_bytes = input_block.bytes();
if (UNLIKELY(input_rows == 0)) {
@@ -504,8 +505,10 @@ Status VTabletWriterV2::write(RuntimeState* state, Block&
input_block) {
// the real 'num_rows_load_total' will be set when sink being closed.
_state->update_num_rows_load_total(input_rows);
_state->update_num_bytes_load_total(input_bytes);
- DorisMetrics::instance()->load_rows->increment(input_rows);
- DorisMetrics::instance()->load_bytes->increment(input_bytes);
+ if (!is_replaying_batched_block) {
+ DorisMetrics::instance()->load_rows->increment(input_rows);
+ DorisMetrics::instance()->load_bytes->increment(input_bytes);
+ }
SCOPED_RAW_TIMER(&_send_data_ns);
// This is just for passing compilation.
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
