This is an automated email from the ASF dual-hosted git repository.
morningman pushed a commit to branch master
in repository https://gitbox.apache.org/repos/asf/doris-website.git
The following commit(s) were added to refs/heads/master by this push:
new b247e512e83 weekly report 260914 (#4164)
b247e512e83 is described below
commit b247e512e83b7fd6c8379a4a7fa1f41531e73a68
Author: Mingyu Chen (Rayner) <[email protected]>
AuthorDate: Mon Sep 21 21:02:46 2026 +0800
weekly report 260914 (#4164)
Co-authored-by: morningman <[email protected]>
---
src/pages/community-report/_reports/2026-09-14.mdx | 473 +++++++++++++++++++++
1 file changed, 473 insertions(+)
diff --git a/src/pages/community-report/_reports/2026-09-14.mdx
b/src/pages/community-report/_reports/2026-09-14.mdx
new file mode 100644
index 00000000000..9522218475f
--- /dev/null
+++ b/src/pages/community-report/_reports/2026-09-14.mdx
@@ -0,0 +1,473 @@
+{/* Auto-generated by community-radar `report export-mdx` — do not edit by
hand.
+ `report` is English-only structured data laid out by doris-website. */}
+
+export const label = "Sep 14 – Sep 20, 2026";
+export const week = "Week 38, 2026";
+export const stats = [
+ {
+ "num": "144",
+ "label": "Merged PRs"
+ },
+ {
+ "num": "23",
+ "label": "New issues"
+ },
+ {
+ "num": "48",
+ "label": "Contributors"
+ }
+];
+
+export const report = {
+ "summary": {
+ "lead": "This week 48 contributors merged 144 PRs, with kernel engineering
remaining the dominant focus, closely followed by compute-storage separation
and real-time data warehouse work—signaling Doris's steady march toward a more
elastic cloud-native architecture and stronger real-time processing. Progress
on lakehouse format interoperability, full-text search acceleration, and
enterprise security governance advanced in parallel, while emerging momentum in
security governance and ag [...]
+ "highlights": [
+ {
+ "title": "Lakehouse Interop: Native Variant, Fluss Access, Iceberg
REST Planning",
+ "narrative": "Doris speeds up unshredded Parquet Variant extraction
and adds native Parquet VARIANT export/import, avoiding JSON re-encoding while
staying cross-engine readable. A new Fluss catalog reads log, primary-key, and
lakehouse-tiered tables; Iceberg REST server-side scan planning advances toward
unified multi-format lakehouse access.",
+ "prs": [
+ {
+ "num": 68036,
+ "title": "[Feature] Support Iceberg REST server-side scan
planning",
+ "url": "https://github.com/apache/doris/issues/68036";
+ },
+ {
+ "num": 67016,
+ "title": "[improvement](be) Speed up unshredded Parquet Variant
extraction",
+ "url": "https://github.com/apache/doris/pull/67016";
+ },
+ {
+ "num": 66399,
+ "title": "[feat](fluss) Support reading Apache Fluss tables
through a fluss catalog",
+ "url": "https://github.com/apache/doris/pull/66399";
+ },
+ {
+ "num": 67424,
+ "title": "[feature](variant) Export and import Variant as the
Parquet VARIANT logical type",
+ "url": "https://github.com/apache/doris/pull/67424";
+ }
+ ]
+ },
+ {
+ "title": "Enterprise Security: Account Locking, Dual-Password
Rotation, Proxy Auth & End-to-End TLS",
+ "narrative": "Doris adds durable MySQL-compatible account locking and
dual-password rotation for zero-downtime credential changes, plus SU...WITH
ROLES proxy auth so gateway tiers run queries as the actual end user with
proper audit. Internal paths now honor TLS end-to-end, and AI resource creation
is restricted by identity—closing governance gaps for multi-tenant,
compliance-driven",
+ "prs": [
+ {
+ "num": 67792,
+ "title": "[feature](auth) Durable ACCOUNT_LOCK / ACCOUNT_UNLOCK
for user accounts",
+ "url": "https://github.com/apache/doris/pull/67792";
+ },
+ {
+ "num": 66115,
+ "title": "[feat](auth) Support MySQL-compatible dual password
(RETAIN CURRENT PASSWORD / DISCARD OLD PASSWORD)",
+ "url": "https://github.com/apache/doris/pull/66115";
+ },
+ {
+ "num": 67672,
+ "title": "[feature](auth) SU ... WITH ROLES: proxy authentication
with mandatory role narrowing, gated by PROXY_PRIV",
+ "url": "https://github.com/apache/doris/pull/67672";
+ },
+ {
+ "num": 68054,
+ "title": "[feature](tls) Align internal communication paths with
TLS",
+ "url": "https://github.com/apache/doris/pull/68054";
+ },
+ {
+ "num": 68147,
+ "title": "[Enhance](resource) Restrict AI resource creation by
user identity",
+ "url": "https://github.com/apache/doris/pull/68147";
+ }
+ ]
+ },
+ {
+ "title": "Full-Text Search Acceleration: Phrase Pushdown, Approximate
LIKE/REGEXP Index, Segment-Aligned Scans",
+ "narrative": "Inverted-index search gets configurable ngram sizing,
candidate-bitmap pushdown into phrase queries, and segment-aligned scan
splitting to cut redundant posting walks. A new approximate gram index
accelerates LIKE/REGEXP pruning, while index_disk_usage exposes index storage
composition—sharpening speed and observability for log and text search
workloads.",
+ "prs": [
+ {
+ "num": 67917,
+ "title": "[feat](inverted-index) Support configurable ngram size
difference",
+ "url": "https://github.com/apache/doris/pull/67917";
+ },
+ {
+ "num": 67180,
+ "title": "[improvement](inverted index) Push the candidate row
bitmap down into phrase queries",
+ "url": "https://github.com/apache/doris/pull/67180";
+ },
+ {
+ "num": 67538,
+ "title": "[feature](inverted index) approximate gram index for
LIKE/REGEXP push-down via ngram tokenizer mode=sparse|dense on SNII",
+ "url": "https://github.com/apache/doris/pull/67538";
+ },
+ {
+ "num": 67915,
+ "title": "[improvement](scan) Align scanner split to segments for
MATCH",
+ "url": "https://github.com/apache/doris/pull/67915";
+ },
+ {
+ "num": 67977,
+ "title": "[feature](inverted index) Add index_disk_usage table
function",
+ "url": "https://github.com/apache/doris/pull/67977";
+ }
+ ]
+ },
+ {
+ "title": "Cloud-Native Elasticity: I/O Coalescing/Read-Ahead,
Object-Storage Spill, Native Azure Access",
+ "narrative": "Doris adds PageIO read-ahead, cross-range I/O
coalescing, and cache-block hole-filling to cut remote I/O dependencies in
compute-storage separation. Cloud mode can now spill to S3 object storage with
billing visibility in SHOW DATA, plus native Azure credential support beyond
the AWS-shim—together sharpening elasticity and cost control.",
+ "prs": [
+ {
+ "num": 67292,
+ "title": "[feature](be) Add PageIO read-ahead, I/O coalescing, and
cache block hole filling",
+ "url": "https://github.com/apache/doris/pull/67292";
+ },
+ {
+ "num": 67612,
+ "title": "[feature](be) Add file-range I/O coalescing and
asynchronous read infrastructure",
+ "url": "https://github.com/apache/doris/pull/67612";
+ },
+ {
+ "num": 68032,
+ "title": "[feature](cloud) Spill to object storage in cloud mode
and report the traffic in SHOW DATA",
+ "url": "https://github.com/apache/doris/pull/68032";
+ },
+ {
+ "num": 68117,
+ "title": "[feature](be) Native Azure credentials, SAS expiry and
client lifetime in S3ClientFactory",
+ "url": "https://github.com/apache/doris/pull/68117";
+ },
+ {
+ "num": 67502,
+ "title": "[feat](cloud) Expose approximate rowset metric",
+ "url": "https://github.com/apache/doris/pull/67502";
+ },
+ {
+ "num": 67621,
+ "title": "[improvement](tablet) Source active tablet stats from BE
reports",
+ "url": "https://github.com/apache/doris/pull/67621";
+ }
+ ]
+ },
+ {
+ "title": "Row Binlog & CDC Maturity: TTL Retention, TSO Pruning,
Stable Column Mapping",
+ "narrative": "Doris adds row/window TTL for ROW Binlog with TSO-based
rowset pruning ahead of scans, cutting initialization cost for bounded
incremental queries. FE-computed stable column mappings decouple read/write
from physical layout, while CDC jobs gain regex table matching, comment sync,
and one-time S3 ingestion—maturing the real-time pipeline end to end.",
+ "prs": [
+ {
+ "num": 68225,
+ "title": "[Feature] Support TTL-based retention and garbage
collection for ROW Binlog",
+ "url": "https://github.com/apache/doris/issues/68225";
+ },
+ {
+ "num": 65858,
+ "title": "[feature](storage) Support row-level TTL and row-binlog
TTL",
+ "url": "https://github.com/apache/doris/pull/65858";
+ },
+ {
+ "num": 68050,
+ "title": "[improvement](be) Prune rowsets by TSO before row binlog
scans",
+ "url": "https://github.com/apache/doris/pull/68050";
+ },
+ {
+ "num": 67456,
+ "title": "[feature](row-binlog) Downlink explicit column mappings",
+ "url": "https://github.com/apache/doris/pull/67456";
+ },
+ {
+ "num": 67974,
+ "title": "[feature](fe) Support regex in CDC
include_tables/exclude_tables",
+ "url": "https://github.com/apache/doris/pull/67974";
+ },
+ {
+ "num": 67975,
+ "title": "[feature](fe) Sync source table comment in CDC streaming
job target tables",
+ "url": "https://github.com/apache/doris/pull/67975";
+ },
+ {
+ "num": 68007,
+ "title": "[improvement](streaming) Support one-time S3 streaming
ingestion",
+ "url": "https://github.com/apache/doris/pull/68007";
+ }
+ ]
+ }
+ ],
+ "numbers": {
+ "mergedPrs": 144,
+ "newIssues": 23,
+ "contributors": 48
+ }
+ },
+ "repos": [
+ {
+ "repo": "apache/doris",
+ "scenarios": [
+ {
+ "name": "Multi-modal Lakehouse",
+ "mergedNarrative": "Lakehouse federation correctness improved across
Iceberg, Paimon, and Hive: schema-evolution-aware equality deletes, safer
name-mapping validation, missing-file handling, and identity-partition
preservation for Iceberg; schema/timestamp precision fixes for Paimon; and
OpenCSVSerde-compatible parsing for Hive CSV tables, reducing silent data
corruption and inconsistent query results across external catalogs.",
+ "merged": [
+ {
+ "num": 67687,
+ "title": "[fix](iceberg) Fix historical scans after schema
evolution",
+ "url": "https://github.com/apache/doris/pull/67687";
+ },
+ {
+ "num": 67904,
+ "title": "[fix](paimon) Preserve statement schemas and timestamp
precision",
+ "url": "https://github.com/apache/doris/pull/67904";
+ },
+ {
+ "num": 68004,
+ "title": "[fix](iceberg) Reject invalid name mapping instead of
reading NULL",
+ "url": "https://github.com/apache/doris/pull/68004";
+ },
+ {
+ "num": 68024,
+ "title": "[fix](be) Truncate native ORC timestamps to
microseconds",
+ "url": "https://github.com/apache/doris/pull/68024";
+ },
+ {
+ "num": 68027,
+ "title": "[fix](hive) Match OpenCSVSerde semantics in
FileScannerV2",
+ "url": "https://github.com/apache/doris/pull/68027";
+ }
+ ],
+ "inProgressNarrative": "Multi-modal lakehouse work fixes native
Azure credential routing for Iceberg files, plus newly opened fixes for
Iceberg/Parquet predicate pushdown mis-pruning matching rows due to NaN
handling and legacy timestamp statistics—preventing silent data loss from
inaccurate metadata—alongside AI resource-related fixes, improving federated
lakehouse query correctness.",
+ "inProgress": []
+ },
+ {
+ "name": "Real-time Data Warehouse",
+ "mergedNarrative": "Real-time analytics got faster planning: eager
aggregation now uses statistics-based greedy join reordering for better
pushdown positions, planner overhead was cut by deferring SQL-cache rendering
and pruning redundant Cascades cost work, partial-update conflict resolution
picks row- vs column-store reads by update mode, and MV rewrite correctly
compensates invalidated roll-up partitions.",
+ "merged": [
+ {
+ "num": 67067,
+ "title": "[improvement](eager-agg) Improve eager aggregation
with greedy join reorder",
+ "url": "https://github.com/apache/doris/pull/67067";
+ },
+ {
+ "num": 67295,
+ "title": "[opt](partial update) Select publish-conflict read
strategy by update mode",
+ "url": "https://github.com/apache/doris/pull/67295";
+ },
+ {
+ "num": 67797,
+ "title": "[improvement](planner) Reduce planner overhead",
+ "url": "https://github.com/apache/doris/pull/67797";
+ }
+ ],
+ "inProgressNarrative": "Real-time warehouse work advances several
key efforts: RPC micro-batching for high-concurrency point queries (highlight)
boosts throughput, MTMV cache moves to a global LRU manager to curb FE heap
bloat from large-scale materialized views, nullable comparison predicates gain
SIMD vectorization, and HBO struct-info work continues (all key-level). New
fixes this week",
+ "inProgress": [
+ {
+ "num": 66688,
+ "title": "[opt](expr) vectorize nullable comparison predicates",
+ "url": "https://github.com/apache/doris/pull/66688";
+ },
+ {
+ "num": 67549,
+ "title": "Hbo structinfo",
+ "url": "https://github.com/apache/doris/pull/67549";
+ },
+ {
+ "num": 67870,
+ "title": "[feature](perf)Support RPC micro-batching for
high-concurrency point queries",
+ "url": "https://github.com/apache/doris/pull/67870";
+ },
+ {
+ "num": 68141,
+ "title": "[opt](mtmv) manage MTMVCache with a global LRU
manager",
+ "url": "https://github.com/apache/doris/pull/68141";,
+ "isNew": true
+ }
+ ]
+ },
+ {
+ "name": "Compute-Storage Separation & Cloud-Native",
+ "mergedNarrative": "Compute-storage separation work improved
efficiency and correctness in write and cache paths: delete-bitmap removal
switched to point deletes to ease FDB pressure, file-cache eviction/TTL
accounting was corrected, commit/version caches and binlog TSO tracking were
hardened against retries and failover, and warm-up jobs now recover cleanly
after FE restarts.",
+ "merged": [
+ {
+ "num": 66598,
+ "title": "[improvement](cloud) Remove pre-rowset delete bitmaps
by key",
+ "url": "https://github.com/apache/doris/pull/66598";
+ },
+ {
+ "num": 67313,
+ "title": "[fix](file cache) keep disk resource limit mode
hysteresis across checks",
+ "url": "https://github.com/apache/doris/pull/67313";
+ },
+ {
+ "num": 67317,
+ "title": "[fix](file cache) count evicted bytes only for blocks
that were downloaded",
+ "url": "https://github.com/apache/doris/pull/67317";
+ },
+ {
+ "num": 67347,
+ "title": "[fix](cloud) bind the packed slice location lifetime
to its writer",
+ "url": "https://github.com/apache/doris/pull/67347";
+ },
+ {
+ "num": 67494,
+ "title": "[fix](cloud) modify
delete_bitmap_store_v2_max_bytes_in_fdb",
+ "url": "https://github.com/apache/doris/pull/67494";
+ }
+ ],
+ "inProgressNarrative": "Compute-storage separation work this week
focuses on metadata consistency and race conditions: new V2 delete bitmap
consistency checks, a refactored packed-file recycling checker, plus fixes for
compute-group property sync, group-recreation mapping, stale commit/rowset
races, delete-bitmap reads across expired transactions, and compaction
overwriting newer bitmaps—most opened this week, hardening correctness between",
+ "inProgress": []
+ },
+ {
+ "name": "Agent Observability",
+ "mergedNarrative": "Variant V2 storage is now the default engine,
with sub-column path pruning fixed for nested element_at chains. Inverted-index
writers stop leaving orphan empty index files, and a new norms-skipping option
shrinks index size for wide Variant schemas; ngram search now validates its
parameters upfront.",
+ "merged": [
+ {
+ "num": 66858,
+ "title": "[feature](variant) Enable Variant V2 by default",
+ "url": "https://github.com/apache/doris/pull/66858";
+ },
+ {
+ "num": 67749,
+ "title": "[fix](search) Validate ngram search gram number",
+ "url": "https://github.com/apache/doris/pull/67749";
+ },
+ {
+ "num": 67859,
+ "title": "[fix](inverted index) Persist empty index files and
stop writing orphan ones",
+ "url": "https://github.com/apache/doris/pull/67859";
+ },
+ {
+ "num": 67955,
+ "title": "[fix](variant) Keep integer element_at indexes out of
variant sub-column paths",
+ "url": "https://github.com/apache/doris/pull/67955";
+ }
+ ],
+ "inProgressNarrative": "This week's key advance connects Variant
values to relational operations: a shared canonical comparator aligns FE/BE
equality semantics, enabling `=`, `!=`, `<=>` comparisons between Variant
expressions (including against NULL) without explicit CAST, and unlocking
Variant hash joins—semi/anti, mark joins, and IN/NOT IN subqueries—for more
natural semi-structured SQL.",
+ "inProgress": [
+ {
+ "num": 67675,
+ "title": "[feature](variant) Support relational operations for
Variant",
+ "url": "https://github.com/apache/doris/pull/67675";
+ }
+ ]
+ },
+ {
+ "name": "Ecosystem Integration",
+ "mergedNarrative": "Cross-engine connectivity improved: Trino
federation avoids invalid predicate pushdown after type casts, JDBC catalogs
correctly resolve SQL Server alias types, and binlog compaction/DELETE handling
was fixed for downstream consumers. Notably, Arrow Flight SQL now serves
session-level SET/GET actions, letting ADBC clients manage catalog, schema, and
variables without raw SQL.",
+ "merged": [
+ {
+ "num": 67966,
+ "title": "[feature](arrow-flight) Serve the Flight SQL session
actions on the Doris session",
+ "url": "https://github.com/apache/doris/pull/67966";
+ },
+ {
+ "num": 67209,
+ "title": "[fix](fe) Prevent invalid Trino string predicate
pushdown",
+ "url": "https://github.com/apache/doris/pull/67209";
+ },
+ {
+ "num": 67916,
+ "title": "[fix](jdbc) Resolve SQL Server user-defined alias
types by JDBC type code",
+ "url": "https://github.com/apache/doris/pull/67916";
+ },
+ {
+ "num": 68033,
+ "title": "[fix](binlog) Preserve row binlog compaction policy",
+ "url": "https://github.com/apache/doris/pull/68033";
+ },
+ {
+ "num": 68034,
+ "title": "[fix](binlog) Fix missing DELETE events in row binlog",
+ "url": "https://github.com/apache/doris/pull/68034";
+ }
+ ],
+ "inProgressNarrative": "Ecosystem compatibility saw several new
fixes this week: JDBC catalogs now honor user-specified character encoding,
MySQL result-set field lengths and prepared-statement parameter forwarding are
corrected, Arrow Flight SQL timezone labels are made portable, CDC streaming
jobs gain better error reporting and manual recovery, and Oracle
BINARY_FLOAT/DOUBLE mapping is added—broadening interoperability",
+ "inProgress": []
+ },
+ {
+ "name": "Security & Governance",
+ "mergedNarrative": "Access-control gaps were closed: row policies
now hold under leading-hint join rebuilds and MOW time-travel queries, binlog
table functions require SELECT privilege, audit redaction no longer leaks
SET_VAR session state, CDC operations run under job creator's identity, and
JDBC driver-URL validation is unified across JDBC, Iceberg, and Paimon
catalogs.",
+ "merged": [
+ {
+ "num": 67776,
+ "title": "[fix](fe) Fix row policy bypass when leading hint
rebuilds the join",
+ "url": "https://github.com/apache/doris/pull/67776";
+ },
+ {
+ "num": 67783,
+ "title": "[fix](policy) Enforce row policies on MOW time travel",
+ "url": "https://github.com/apache/doris/pull/67783";
+ },
+ {
+ "num": 67786,
+ "title": "[fix](audit) Fix SET_VAR leakage after INSERT audit
logging",
+ "url": "https://github.com/apache/doris/pull/67786";
+ },
+ {
+ "num": 67851,
+ "title": "[fix](cdc) Use streaming job creator identity for CDC
operations",
+ "url": "https://github.com/apache/doris/pull/67851";
+ },
+ {
+ "num": 68088,
+ "title": "[fix](binlog) Require SELECT privilege for binlog TVF",
+ "url": "https://github.com/apache/doris/pull/68088";
+ }
+ ],
+ "inProgressNarrative": "Security governance adds a new fix this week
rejecting name collisions between tables and dictionaries in the same
database—since they share a privilege namespace but previously weren't
cross-checked, risking ambiguous privilege keys—alongside identifier quoting to
prevent injection in dictionary loads and stricter authorization-context guards
for reused prepared point-query plans.",
+ "inProgress": []
+ },
+ {
+ "name": "Kernel & Engineering",
+ "mergedNarrative": "Kernel engineering delivered broad correctness
and performance hardening: fixes across the optimizer (union, TopN pushdown,
aggregate rewrites), scalar functions (date, string, IP, decimal casts),
zonemaps, window functions, and compaction, plus refactors isolating BE Java
plugins, extending key-range compaction, and removing legacy
minidump/native-file-format code.",
+ "merged": [
+ {
+ "num": 66729,
+ "title": "[refactor](be-java-extensions) isolate every BE Java
extension into its own plugin",
+ "url": "https://github.com/apache/doris/pull/66729";
+ },
+ {
+ "num": 67177,
+ "title": "[fix](local shuffle) Prevent row loss at
parallel-to-serial pipeline boundaries",
+ "url": "https://github.com/apache/doris/pull/67177";
+ },
+ {
+ "num": 67276,
+ "title": "[fix](be) Fix SIGSEGV in bvar::take_sample caused by
~AgentCombiner walking freed TLS agents",
+ "url": "https://github.com/apache/doris/pull/67276";
+ },
+ {
+ "num": 67360,
+ "title": "[improvement](compaction) Support key ranges in rowset
merger",
+ "url": "https://github.com/apache/doris/pull/67360";
+ }
+ ],
+ "inProgressNarrative": "Kernel engineering is the largest group,
highlighted by a native UUID data type spanning FE/BE/storage/expressions. Most
remaining items, largely opened this week, are correctness fixes across
arithmetic simplification, CASE/NULL semantics, TopN lazy materialization,
hidden-column (TSO/VERSION) read consistency, ARM64 compatibility,
JSONB/IPv6/timestamp edge cases, and query-cancellation resource leaks—steadily
hardening the execution engine",
+ "inProgress": [
+ {
+ "num": 67627,
+ "title": "[feature](datatype) Add UUID data type support",
+ "url": "https://github.com/apache/doris/pull/67627";
+ }
+ ]
+ }
+ ],
+ "demand": [
+ {
+ "name": "Multi-modal Lakehouse",
+ "narrative": "The community wants Doris to fully support Iceberg
REST server-side scan planning. Currently, despite adopting a newer Iceberg
Java SDK, Doris downgrades RESTTable into a plain BaseTable, losing planning
behavior and bypassing safeguards for unsupported modes. This work aims to
properly retrieve plans via the SDK and execute server-returned tasks,
improving Iceberg REST Catalog compatibility and efficiency.",
+ "refs": [
+ {
+ "num": 68036,
+ "title": "[Feature] Support Iceberg REST server-side scan
planning",
+ "url": "https://github.com/apache/doris/issues/68036";
+ }
+ ]
+ },
+ {
+ "name": "Real-time Data Warehouse",
+ "narrative": "Since ROW Binlog is retained indefinitely, creating
storage and maintenance burdens, the community proposes a TTL-based retention
and garbage collection policy. This end-to-end design covers query visibility,
background discovery of expired data, compaction, and file reclamation across
both shared-nothing and cloud-native deployments, better supporting incremental
queries, table streams, and materialized view maintenance.",
+ "refs": [
+ {
+ "num": 68225,
+ "title": "[Feature] Support TTL-based retention and garbage
collection for ROW Binlog",
+ "url": "https://github.com/apache/doris/issues/68225";
+ }
+ ]
+ }
+ ]
+ }
+ ]
+};
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
