This is an automated email from the ASF dual-hosted git repository.
hanahmily pushed a commit to branch vectorized-query
in repository https://gitbox.apache.org/repos/asf/skywalking-banyandb.git
The following commit(s) were added to refs/heads/vectorized-query by this push:
new 91460b3b8 docs(CHANGES): reflect G9a-G9e single-node vectorized
coverage complete
91460b3b8 is described below
commit 91460b3b831f2662dc7b4bddd861793b6dbd141c
Author: Hongtao Gao <[email protected]>
AuthorDate: Sat May 16 02:34:18 2026 +0000
docs(CHANGES): reflect G9a-G9e single-node vectorized coverage complete
The 0.11.0 vectorized bullet still read as if only scan + GroupBy+Agg
(+ order_by) were vectorized and that TopN / hidden-criteria-tag
queries flowed through the row path. G9a-G9e closed every single-node
fall-through: TopN/BottomN, scalar reduce, raw GroupBy, projection
auto-coverage, hidden criteria tags, and boundary-error parity now all
resolve through the vec dispatch, soak- and bench-validated.
Rewrite the release note to state single-node coverage is complete and
scope the remaining row-path fall-throughs precisely to the distributed
Map-mode / multi-measure / non-vec-backend cases (future distributed
vectorized query work). No code change.
---
CHANGES.md | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/CHANGES.md b/CHANGES.md
index 49b87c729..9c7e52d4e 100644
--- a/CHANGES.md
+++ b/CHANGES.md
@@ -5,7 +5,7 @@ Release Notes.
## 0.11.0
### Features
-- Vectorized measure query path is now enabled by default. The columnar
pipeline replaces per-row protobuf serialization in `NewMIterator`, cutting
allocations and ns/op for scan-heavy measure queries; gRPC wire format
(`*measurev1.InternalDataPoint`) is byte-identical. Coverage extends to
single-node GroupBy+Agg via `BatchAggregation` (SUM/COUNT/MIN/MAX/MEAN with
row-path-equivalent type semantics and first-seen carry-forward of non-key
projected tags). Distributed Map-mode GroupBy+Agg, [...]
+- Vectorized measure query path is now enabled by default. The columnar
pipeline replaces per-row protobuf serialization in `NewMIterator`, cutting
allocations and ns/op for scan-heavy measure queries; gRPC wire format
(`*measurev1.InternalDataPoint`) is byte-identical. Single-node coverage is
complete: scan, GroupBy+Agg via `BatchAggregation`, scalar reduce (`Agg`
without `GroupBy`), raw `GroupBy` (without `Agg`), implicit projection coverage
for GroupBy/Agg fields, `TopN`/`BottomN`, `o [...]
- Add validation to ensure Measure's ShardingKey contains all Entity tags to
guarantee entity locality.
- Organize access logs under a dedicated "accesslog" subdirectory to improve
log organization and separation from other application data.
- Collect BanyanDB data on e2e test failure for CI debugging.
