iceberg+starrocks=🔥 StarRocks 4.1 now supports native SQL

DELETE

for Apache Iceberg tables — no Spark jobs, custom scripts, or external rewrite pipelines required. 👉 Read the full blog to see how it works and why it matters: https://medium.com/starrocks-engineering/starrocks-4-1-no-more-external-rewrite-pipelines-for-apache-iceberg-deletes-ce13132a082e?postPublishedType=repub

Release - 3.5.17 New release published by yingtingdong Release date: May 13, 2026 Behavior Changes •

SHOW CREATE TABLE

and

DESC

now show Primary Keys for Paimon tables. #70535 • Disallowed INSERT into insert-only ACID Hive tables in Hive catalogs. #71460 •

START_TIME

and

END_TIME

in Profile are now displayed using the session time zone. #71429 Improvements • Supports

csv.enclose

and

csv.escape

in

INSERT INTO FILES

CSV export. #71589 • Added query relation information (directly queried tables and viewa) to audit logs. #71596 • Made the FE configuration

star_mgr_meta_sync_interval_sec

runtime mutable. #71675 • Reduced metadata and lock overhead in table metadata and row-count paths. #72053 #72042 #71672 • Improved build and dependency hygiene by merging the broker builder into the FE build and removing WildFly OpenSSL. #71823 #71908 Bug fixes The following issues have been fixed: • Wrong results for local-shuffle aggregate queries with OFFSET. #71997 • Incorrect Join output properties after Exchange shuffle columns are pruned. #72003 • Several dependency CVE issues. #71762 #71914 • Oracle JDBC NLS format handling issue. #71412 • Missing Iceberg column statistics in manifest data file cache. #71913 • Missing Hive partition directory before INSERT OVERWRITE commit. #71810 • Aggregate-join-pushdown materialized view rewrite and min/max optimization issues on Iceberg base tables. #71856 #71863 • Race conditions in

ConnectorSinkPassthroughExchanger

and

LoadChannel::get_load_replica_status

. #71848 #71843 • Credential redaction issue in INSERT FILES operations. #71245 • Incorrect

reverse(DecimalV3)

results. #71834 • Missing JNI exception handling checks in Java UDF code. #71734 • Incorrect short-circuit checks in

EventScheduler

. #71740 • Incorrect Arrow Flight column name for empty result sets. #71534 • Batch publish deadlock caused by partition version gaps. #71483 • Repeated Apply attachments in scalar-subquery plans. #71155 StarRocks/starrocks

Hi community, 👋 Recently we’ve published two new blogs on starrocks*StarRocks 4.1 + Apache Iceberg*iceberg: 1️⃣ StarRocks 4.1: Native Support for Semi-Structured Data in Iceberg Learn how StarRocks 4.1 brings Iceberg Variant into distributed execution, enabling faster queries on semi-structured data without repeated JSON parsing. 👉 https://medium.com/starrocks-engineering/starrocks-4-1-native-support-for-semi-structured-data-in-iceberg-6c71e989dd63 2️⃣ StarRocks 4.1: Iceberg Acceleration That Stays Fresh Explore how Incremental Materialized Views on Iceberg help keep acceleration fresh by refreshing based on data changes instead of full table history. 👉 https://medium.com/starrocks-engineering/starrocks-4-1-iceberg-acceleration-that-stays-fresh-9bc962c056a4 If you’re working with Iceberg, semi-structured data, or large-scale lakehouse analytics, we’d love for you to give them a read and share your thoughts in the thread. 👇

Release - 4.0.11 New release published by wangsimo0 Release Date: June 5, 2026 Behavior Changes •

get_json_string

and the other

get_json_*

functions now return the JSON parse error instead of NULL when implicit VARCHAR-to-JSON parsing fails under

ALLOW_THROW_EXCEPTION

. The default behavior (returning NULL when the mode is disabled) is unchanged. #73199 •

pipeline_enable_large_column_checker

is now enabled by default. #72798 Improvements • Lake write-path load spill files now use a flat, single-level directory layout with the transaction ID baked into each filename, and are reclaimed by a txn-id-based vacuum pass. This moves bulk deletes off the write hot path and lets vacuum clean up spill files leaked by BE crashes. #73064 • SHOW statements (such as

SHOW GRANTS

and

SHOW WAREHOUSES

) are now allowed inside an explicit transaction, so BI/JDBC clients that automatically issue SHOW no longer break the transaction flow. #72954 • Java UDAF and UDTF now support STRUCT arguments and return types. #72911 • Scalar Java UDF now supports STRUCT arguments. #72620 • Java UDF now supports DATE and DATETIME types. #72337 • Java UDF now supports nested ARRAY/MAP types. #72283 • Added the FE configuration

deploy_serialization_min_thread_pool_size

. #72274 • Skipped redundant partition key expression building when an

add_partition_value

deduplication hit occurs. #73156 • Avoided a redundant

latestSnapshot()

call in

PaimonMetadata#getTableVersionRange

. #72892 • Deduplicated commutative AND/OR expressions in scalar operator common subexpression elimination. #72823 Bug Fixes The following issues have been fixed: • A memory leak introduced by the UDAF cache. #74025 • An incorrect implementation in aggregate combined functions. #74169 • An issue in shared-data combined txn log mode where the per-partition coordinator claim was not re-recorded on every sender's open, which could drop txn logs. #73962 • A read failure on Iceberg tables that use a custom

LocationProvider

, fixed by lazily initializing the

LocationProvider

in

SerializableTable

. #73482 • A serialization failure caused by the

de.javakaffee

UnmodifiableCollectionsSerializer

, now replaced with a Java 17-compatible version. #73458 •

HdfsFsManager

copy error messages now include the underlying cause. #73414 • A concurrent

SegmentFlushTask

race in

DeltaWriter::commit()

. #73371 • Sort merge provider errors are now propagated to the fragment context instead of being lost. #73337 • An issue where Ranger row-filter/masking policies on Hive views were skipped, so policies on the view or its base tables were not applied. #73265 • Upgraded libthrift to 0.23.0 to address a security vulnerability (CVE). #73243 • An FE file-descriptor leak, fixed by reusing

HttpClient

instances. #73239 • Parquet broker load errors now include file/column/row context. #73236 • A slot lookup failure for output slots with an empty

col_name

in the Spark connector external scan. #73225 • A crash in

SinkBuffer

during graceful exit. #73202 • Query cache conflicts with local shuffle aggregation. #73194 • A use-after-free of the Hive partition descriptor across fragment teardown. #73176 • A thread-safety issue in lake vacuum, fixed by using

localtime_r

. #73088 • A race condition between

PipelineTimerTask

doRun

and unscheduling during query context destruction. #73082 • Lock contention on read-only query-engine paths, reduced by relaxing DB locks. #73067 • An materialized view refresh failure with SQL Server tables in a JDBC catalog. #72962 • A JNI local-reference leak in

JDBCScanner::_init_jdbc_scanner

. #72913 • An issue where partition TopN could lose a child's output column. #72848 • An incorrect plan caused by not clearing

LambdaArgument.transformedOp

before INSERT OVERWRITE re-planning. #72832 • The coordinator lock was held during external resource cleanup. #72830 •

Locker

rollback is now exception-safe and the unlock order is fixed. #72789 • An incorrect byte order in

ColumnDict.merge

, now using unsigned byte order. #72778 • A stack-buffer-overflow when formatting into a temporary

std::string

. #72728 • The HAVING clause is now checked when disabling aggregation spill on a small LIMIT. #72705 • A hang caused by joining forwarded RPCs when draining the runtime_filter worker. #72626 • Incorrect lazy-materialization slot nullability for a materialized view over an outer join. #72621 •

merge_condition

was not preserved when applying a normal rowset commit. #72542 • Lock contention in

TabletScheduler

/

TabletSchedCtx

hot paths during clone, reduced by relaxing DB locks. #72475 •

Locker

did not roll back a partial intensive-lock acquisition. #72423 • A spillable hash join probe crash. #72397 • COALESCE children are now cast to a common type in the JOIN USING transformer. #72338 • DB READ lock was held too broadly for single-table proc directories, now relaxed to per-table. #72334 • A memory leak when caching the materialized view plan context. #72300 • FSE-v2 did not set the schema for shared-data sorted schema change. #72235 •

ConsistencyChecker

held a DB READ lock too broadly in periodic scans, now relaxed to per-table READ. #72218 • A BE crash when querying

information_schema.warehouse_queries

. #72019 • A trailing

\r

was not stripped before the closing enclose in CRLF … StarRocks/starrocks

Release - 3.5.18 New release published by yingtingdong Release date: June 5, 2026 Behavior Changes •

SHOW

statements are now allowed inside explicit transactions. #72954 •

get_json_string

now respects

ALLOW_THROW_EXCEPTION

when handling JSON parsing errors. #73199 •

IGNORE NULLS

is now preserved in view definitions when the window function argument is an expression. #69971 • Ranger row filter and masking policies are now correctly applied to Hive views and to base tables expanded from Hive view definitions. #73265 • Hive partition statistics are no longer automatically refreshed per partition. Existing cached stats are preserved while a table-level asynchronous refresh updates the cache in batches. #73563 Improvements • Supports caching Java UDAF class-level initialization so shared UDAFs can reuse loaded classes and generated stubs across aggregator and window-function instances. #72038 • Supports Paimon time types and improves Paimon materialized view handling. #58292 • Added an Avro schema cache for shadowed

PartitionData

during partition load. #72215 • Added a configurable FE write timeout

mysql_send_packet_timeout_ms

for the MySQL result send path to prevent indefinitely blocked result sending to slow clients. #73646 • Optimized

CatalogRecycleBin

adjusted recycle timestamp lookup. #72128 • Reduced metadata and lock overhead in load balancing, compaction scheduling, consistency checks, and StarMgr metadata synchronization paths. #73555 #72218 #72178 #72108 • Improved diagnostics for filesystem copy failures and Parquet broker load errors by surfacing the underlying cause and file/column/row context. #73414 #73236 • Reduced external catalog and information schema metadata overhead by deferring JDBC REMARKS fetching, avoiding redundant Paimon snapshot lookups, and pushing down

table_name

predicates for

information_schema.tables_config

. #73488 #72892 #73210 • Simplified the scalar-function merge implementation by using

merge()

directly. #69575 Bug fixes The following issues have been fixed: • Empty

ALTER TABLE

statements could be parsed as OPTIMIZE clauses, and replaying malformed OPTIMIZE jobs could clear a table's default distribution. #73352 • Decimal-valued unit counters in runtime profiles could cause query progress parsing failures and noisy FE warnings. #73683 • Concurrent

SegmentFlushTask

race in

DeltaWriter::commit()

and loss of

merge_condition

during normal rowset commit. #73371 #72542 • Crashes, hangs, or unsafe cleanup in

SinkBuffer

graceful exit,

PipelineTimerTask

, runtime filter workers, spillable hash join probe,

information_schema.warehouse_queries

, lake vacuum, HTTP connection unregister paths, and query queue timeout handling. #73202 #73082 #72058 #72626 #72397 #72019 #73088 #72006 #65802 • Materialized view issues involving JDBC SQL Server tables, lost index properties, cached plan context memory leaks, Paimon tables, and incorrect shuffle distribution after MV rewrite. #72962 #69187 #72300 #58292 #71075 • Query planning and rewrite issues in Spark connector external scans,

INSERT OVERWRITE

re-planning, aggregation spill with small LIMIT, and generated columns produced by

UNNEST

. #73225 #72832 #72705 #72027 • Paimon Primary Key columns could be incorrectly marked as non-nullable when querying external catalogs. #71660 • Primary Key and tablet metadata issues including partial tablet schema short-key mismatch, rowset metadata cache warmup deadlock, disk data cache expansion failure, Azure filesystem client cache issues in Starlet, and colocate-heavy cluster-balance performance issues in StarOS. #70586 #71459 #58206 #73145 #72391 • Locker rollback and unlock-order issues during partial intensive-lock acquisition. #72789 #72423 • Dependency CVEs and broker dependency regressions. #72905 #72797 #72184 #72191 • JNI local-reference leaks in JDBC scanner initialization. #72913 • Arrow dictionary values in Parquet scanner and Apache Parquet namespace ambiguity during scanner builds. #71855 #72284 • NPE in Iceberg

getPartitionLastUpdatedTime

when the snapshot is expired. #68925 StarRocks/starrocks

Hi Community, 👋 StarRocks now has an integration with OpenMetadata, making it easier to catalog StarRocks metadata and support workflows like metadata ingestion, query usage, lineage, and profiling. cool rocky 🚀 To get started, check out the official OpenMetadata setup guide: https://docs.open-metadata.org/v1.12.x/connectors/database/starrocks ✅ For a hands-on walkthrough with real test results and setup notes, read @김병주’s tested guide here: https://starrockskorea.super.site/starrocks-library/starrocks-with-openmetadatabeta

Friendly reminder: Our webinar, Introducing PhoenixAI: Built for Agents That Query, Reason, and Act, is happening tomorrow at 10:00 AM PT! Save your seat here: https://bit.ly/4eyPQ5Y yay <!channel> AI agents are creating a new set of demands on the data layer — from agent-scale latency and concurrency to ad hoc queries, fresh data, and governance built for autonomous workloads. In this session, our product expert, Sida Shen, will unpack what that means in practice: why existing analytical infrastructure wasn’t designed for these patterns, what PhoenixAI Cloud supports today, and how we’re building toward one analytical database for the full agent loop: decision → execution → trace. We’ll also have live Q&A at the end. If you can’t join live, register anyway and we’ll send the recording afterward! https://bit.ly/4eyPQ5Y

Hi Community, 👋 Have you tried @Rocky in the #C02FACZSNJV channel?rocky nice Next time you have a StarRocks question, give it a try! Today, we’re sharing the story behind Rocky: how we built this RAG assistant, the architecture choices we made, and the lessons we learned along the way. If you’re using StarRocks and thinking about building lightweight AI applications, this post is worth checking out: https://medium.com/dev-genius/lessons-we-learned-building-a-rag-assistant-without-a-separate-vector-database-26df51f33219 We’d love to hear from you: are you using StarRocks to build AI apps? What feedback do you have for improving Rocky? Drop your thoughts in the comments.👇

Release - 4.1.1 New release published by celerdata2308 4.1.1 Release Date: May 29, 2026

Container Image Issue (v4.1.0): Due to an unstable load order issue in the v4.1.0 container image, BE processes may fail to start reliably in container environments. Container environment users should NOT upgrade to v4.1.0; use v4.1.1, which includes the fix (#71825).

Downgrade Notes: After upgrading StarRocks to v4.1, DO NOT downgrade to any v4.0 version below v4.0.6. Due to internal data-layout changes in v4.1 (tablet splitting and distribution), downgrade from v4.1 is only supported to v4.0.6 or later.

Behavior Changes • The Hive connector now uses a native C++ Avro scanner instead of the JNI Avro scanner by default. #73237 #73569 • Query rewrite over INCREMENTAL/AUTO materialized views is now disabled, and FORCE refresh and partition refresh are rejected for INCREMENTAL/AUTO materialized views. #72890 #72336 #71355 Improvements • Java UDF/UDAF/UDTF now support more types: STRUCT arguments and return values for UDAF/UDTF, nested ARRAY/MAP types, DATE/DATETIME, DECIMAL, and varargs. #72911 #72283 #72337 #72208 #68596 • Scalar UDFs now support STRUCT arguments. #72620 • Python UDFs now support nested ARRAY/MAP types. #72210 • UDAFs are now loaded and initialized once and reused across queries, reducing per-query overhead. #72038 • Replaced the JNI Avro scanner with a native C++ scanner for the Hive connector, with direct binary decoding and support for

avro.schema.literal

and

avro.schema.url

. #73237 #73283 #73257 #73569 • Supports the Trino

WITH

clause in CTAS statements. #71960 • Completed Iceberg

timestamptz

partition transform support on the sink path. #73397 • Enabled TopN runtime filter pushdown for Iceberg table aggregation. #72332 • Supports Iceberg datetime min/max optimization. #71870 • Allows HDFS HA configuration passthrough in Catalog and BE to support accessing multiple HDFS clusters. #71521 • Added a partition scan number limit for external table queries. #68480 • Fails fast for unsupported Iceberg V3 features. #70242 • Supports

csv.enclose

and

csv.escape

for CSV exports via INSERT INTO FILES. #71589 • Added the

enable_push_down_schema

INSERT property for full schema push-down to

files()

. #70978 • Routine Load jobs are now paused on non-retryable errors (for example, primary key size exceeded). #71161 • Supports join reorder for complex expressions from two children. #71615 • Improved CBO statistics estimation, including MCV/null-fraction propagation for

date_trunc

,

array_map

, CASE WHEN, IS NULL, UNION, and constants. #72233 #70372 #70221 #70865 #70989 #71000 • Improved skew join detection: skew is only detected when all join keys are skewed, and a

force_group_by_skew_eliminate_when_skewed

switch was added to force the skew rule. #72753 #71382 • Supports constant folding for

regexp_replace

in the FE. #70804 • Optimized MIN/MAX on date partition columns with constant partition values. #69880 • Introduced the

SCHEDULE

keyword as a synonym for

ASYNC

in materialized view refresh. #72329 • Supports tablet creation retry for Lake tables in shared-data mode. #71068 • Supports conditional update for Lake column-mode partial update. #71961 • Parallelized partial-update publish, persistent index initialization, and SSTable opening to improve ingestion throughput. #71652 #71217 #72112 #71145 #72986 • Supports DCG file synchronization during shared-nothing to shared-data replication. #69339 • Supports schema evolution for widening VARCHAR length on both key and non-key columns. #70747 • Added the

snapshot_meta.json

marker for cluster snapshot integrity checks. #71209 • Supports LDAP direct bind authentication via a DN pattern. #71559 • Added the

get_query_dump_from_query_id

meta function for easier query troubleshooting. #72875 • Supports auditing queried relations in the audit log. #71596 • Added session variables for MySQL binary result encoding. #71415 • Added metrics for better observability, including

tablet_num

for shared-data clusters,

MemtableIOSpeed

,

staros_shard_count

, and Iceberg metadata-table query metrics. #71444 #69842 #73096 #70825 • Added the FE configuration

deploy_serialization_min_thread_pool_size

. #72274 • Added the

tablet_reshard_enable_tablet_merge

configuration to disable MergeTabletJob creation. #70906 • Eliminated HTTP-server accept thundering-herd via

SO_REUSEPORT

. #72956 Security • [CVE] Upgraded Netty to 4.1.133.Final. #72905 • [CVE-2026-42198] [CVE-2026-5598] Bumped pgjdbc to 42.7.11 (client-side DoS via unbounded SCRAM PBKDF2 iteration count) and BouncyCastle to 1.84 (FrodoKEM private-key leakage). <https://github.com/StarRocks/starrocks/pull/7279… StarRocks/starrocks

Release - 4.0.12 New release published by wangsimo0 4.0.12 Release Date: June 25, 2026 Behavior Changes • When reading INT64 timestamps from Parquet files written with

isAdjustedToUTC=false

(timezone-naive),

SELECT FROM FILES()

and broker/stream LOAD no longer shift the values by the session timezone offset. Such timestamps are now read as wall-clock values, consistent with Trino, Spark, and Impala. Previously the values drifted whenever the session timezone was not UTC. #73674 • CTAS (

CREATE TABLE AS SELECT

) now preserves the declared

VARCHAR(N)

length when the source carries an explicit user length (a catalog column reference,

CAST AS VARCHAR(N)

, or a string literal), instead of widening it to

VARCHAR(1048576)

. This keeps the length constraint enforceable and aligns DDL with dbt schema contracts. Materialized view materialization still widens columns as before. #73498 • The Paimon connector now respects the session variable

connector_max_split_size

when calculating scan splits, instead of always using the default value, so tuning it now affects Paimon scan parallelism. #71756 Improvements • Optimized

base64_to_bitmap

by folding the conversion at constant-evaluation time for constant inputs. #74684 •

ngram_search

now supports a non-constant needle (the search term can be a column expression rather than only a constant). #74675 • The Arrow-to-JSON converter now supports

LARGE_LIST

and

FIXED_SIZE_LIST

types. #73714 • Added an opt-in option to isolate wide-string columns during statistics collection to reduce memory pressure. #73258 •

information_schema.COLUMNS

now populates the

DATETIME_PRECISION

field. #74623 • Relaxed database read locks to table-scoped intensive locks in

InformationSchemaDataSource

and

FrontendServiceImpl

to improve concurrency. #73936 #73913 • Narrowed database write locks to table-scoped intensive write locks for shared-nothing clusters, and scoped replica row-count updates to the table lock. #74523 #74521 • Moved the routine-load broker RPC out of the per-job write lock to reduce contention. #73591 • Deferred JDBC

REMARKS

fetching out of the

getTable()

hot path to speed up metadata access for JDBC catalogs. #73488 • Pushed down the

table_name

predicate for

information_schema.tables_config

queries. #73210 • Skipped per-replica scans on single-medium BEs in

BackendLoadStatistic

. #73555 • Added a write timeout to the MySQL channel result send path to prevent stuck connections. #73646 • Added catalog recycle bin size gauge metrics. #74440 • Added vacuum batch-size and retry-count metrics, and added decorrelated jitter to the lake vacuum retry backoff to reduce retry storms. #74112 #74108 • Upgraded third-party dependencies to address security vulnerabilities (CVE): Netty to 4.1.135.Final, Tomcat to 9.0.118, and Thrift to 0.23.0. #74668 #73797 #73625 Bug Fixes The following issues have been fixed: • Successfully committed multi-statement transaction stream loads were shown as

PREPARING

forever in

information_schema.loads

and

SHOW STREAM LOAD

. #74386 • Rows were silently dropped from

information_schema.loads

on clusters whose session timezone differs from Asia/Shanghai, because load times were exchanged as naive wall-clock strings across the BE/FE thrift boundary. #73365 • The

COMMIT

of an explicit transaction waited only

query_timeout

milliseconds (instead of seconds) for the database write lock due to a unit mismatch. #73549 •

current_timestamp

/

now()

column defaults were displayed as a frozen literal after

ALTER TABLE ... ADD COLUMN

and could be lost across FE restarts or edit-log replay. #73455 • Querying

sys.fe_memory_usage

/

sys.fe_locks

without the

OPERATE ON SYSTEM

privilege returned a misleading RPC-failure message instead of a clear access-denied error. #73567 • Automatic per-key Hive partition stats refresh could overload the Hive Metastore for tables with many partitions. #73563 • A null-pointer issue when reading the GTID during a schema change. #74855 • An empty analytic operator was not pruned after pushing down a distinct aggregation. #74810 • Zero row counts could corrupt partition statistics. #74801 • Vector index rewrite could pollute the shared table schema. #74785 • An

IllegalStateException

during parallel profile collection, fixed by making Tracers fork-aware. #74746 • BE vacuum tasks were not aborted once the FE caller's timeout elapsed. #74694 • Partition consumer errors in

ChunksPartitioner

were lost instead of being propagated. #74693 • A lock mismatch in

blockingAddTabletCtxToScheduler

. #74596 • A typo in the

azure_adls2_oauth2_client_endpoint

configuration field name. #74581 • Pipeline observers were not notified on missed operator state transitions. #74557 • The reported vacuum watermark was incorrect when retain-boundary metadata was gone. #74429 • A data race on

MaterializedIndexMeta

during

updateSchemaBackendId

. #74412 • A non-primary-key replica could get stuck with a permanent version hole; it now self-heals. #74408 • A use-after-free of

LLVMContext

when JIT compilation fails. #74396 • A column mismatch in the missing-replica row of

ADMIN SHOW REPLICA STATUS

. #74393 • Invalid JIT IR generated for

CASE WHEN

with mixed float/int WHEN and result types. #74382 • The

CatalogRecycleBin

was frozen when a cluster snapshot kept failing. #74379 • A partial update targeting a table modified earlier in the same explicit transaction is now rejected with a clear error. #74344 • Immutable-partition updates did not use the transaction's compute resource. <https://github.c… StarRocks/starrocks

🆕 Integration:starrocks StarRocks x DataHub The latest blog walks through how to make StarRocks metadata discoverable in DataHub, including tables, views, external catalogs, profiling metadata, and column-level view lineage — all in one searchable metadata layer. Check it out here: https://medium.com/starrocks-engineering/make-your-starrocks-metadata-discoverable-with-datahub-f1a033e6f987

Release - 3.5.19 New release published by yingtingdong Release date: June 26, 2026 Behavior Changes •

parse_json

now respects `ALLOW_THROW_EXCEPTION`: malformed JSON fails the query instead of silently producing

NULL

, mirroring the earlier

get_json_string

change. #74976 •

FILES()

and Broker Load now honor the Parquet

isAdjustedToUTC=false

flag for

INT64

timestamps, so such timestamps are no longer shifted by the session time zone. #73674 •

SHOW FUNCTIONS

now surfaces the

isolation

property (

shared

or

isolated

) for Java UDFs and UDAFs. #75255 • The non-reserved keywords

FLOOR

and

CEIL

are now allowed as column names. #75241 Improvements • Supports loading Arrow

LARGE_LIST

and

FIXED_SIZE_LIST

columns into

JSON

columns via

FILES()

and Broker Load. #73714 #73718 • Added an opt-in

statistics_large_string_column_merge_threshold

to isolate wide `CHAR`/`VARCHAR` columns into dedicated statistics collection. #73258 • Optimized

base64_to_bitmap

for constant inputs and hardened it against invalid base64-encoded bitmaps. #74684 • Added new metrics for lake vacuum batch size and retry counts, and gauges for

CatalogRecycleBin

size. #74112 #74440 • Supports auditing a statement twice. #73896 Bug fixes The following issues have been fixed: • Several wrong-result and planning issues: low-cardinality dictionary translation for expressions where

f(NULL)

is not `NULL`; a

Multiple entries with same key

error from common-subexpression elimination of commutative `AND`/`OR`; an

AGGREGATE has mismatch types

error; a compound predicate with an always-false nested branch under

UNION

returning no rows; and an off-by-one that dropped a row in

RANK

TopN at a chunk boundary. #69376 #72823 #74159 #74218 #75045 • Materialized view rewrite issues that could return incorrect results: aggregate MVs with a

HAVING

clause rewriting queries without (or with weaker)

HAVING

, and

avg(DISTINCT x)

being rewritten through a `sum`/`count` MV. #73610 #75071 • Incorrect window-function results when

enable_push_down_pre_agg_with_rank

split a window count into local pre-aggregation and global analytic merge, and an empty window operator generated after pushing down distinct aggregation. #74453 #74810 • Partition TopN losing a child operator's output column, and silently swallowing sort or pre-aggregation errors and returning wrong or partial results. #72848 #74693 • Iceberg equality-delete rows with

NULL

identity-column values were not applied. #67321 • A spurious strict-mode cast overflow error raised from undefined data in

NULL

rows. #74903 • Decimal scale could be lost when a column is entirely

NULL

. #73789 • BE crashes in

to_base64

(stack overflow), JSON load of nested types via partial append, local partition TopN with a non-nullable aggregate result, partitioned join (out-of-bounds from inaccurate memory accounting), runtime profile serialization (counter min/max race), JIT compilation failure (use-after-free of

LLVMContext

), invalid JIT IR for

CASE WHEN

with mixed float/integer types, and partial column updates under schema drift. #70623 #73715 #69752 #74315 #72904 #74396 #74382 #74005 • An out-of-bounds read and potential oversized allocation in

split

,

split_part

, and

str_to_map

when the input ends with a truncated UTF-8 byte. #75068 • A memory leak from the UDAF context cache and inflated query-pool memory accounting in

OlapTableSink

. #74025 #73807 • Unexpected backend process restarts. #74424 • Materialized view issues: a slot-nullability crash for MVs defined with

FULL OUTER JOIN

under late materialization, an NPE refreshing nested MVs, a duplicated warehouse property in

SHOW CREATE MATERIALIZED VIEW

, and a vector ANN query polluting a shared table schema and breaking unrelated statements. #72621 #73644 #69418 #74785 • Querying Paimon tables whose

DATE

partition column contains

NULL

values. #73950 • Reading Hudi MOR tables with `char`/`varchar` columns when

hudi_mor_force_jni_reader

is enabled. #58521 • Nested

INT96

timestamps (inside

ARRAY

,

MAP

, or

STRUCT

) were shifted by the session time zone during `FILES()`/Broker Load. #74868 • Incorrect bytes-read statistics in the audit log for connector scans, and incremental connector scan ranges being assigned to driver sequences absent from the deployed fragment, which could drop part of the scan. #73799 #74674 • Meta scan could fail after schema changes such as

ADD COLUMN

, which could fail background statistics collection. #72901 • Slow broker RPCs held the per-job Routine Load write lock and blocked admin RPCs and

SHOW ROUTINE LOAD

. #73591 •

ALTER ROUTINE LOAD

persisted an invalid statement for reserved-keyword table names, which could drop the load clause on FE restart. #74188 • `GRANT`/`REVOKE` on the

public

role did not invalidate cached merged privileges, leaving stale authorization. #73717 • A race allowing concurrent operations to observe torn state during table and materialized-view

RENAME

and

SWAP

, and a data race on

MaterializedIndexMeta

schema-update tracking. #74100 #74412 • Database-level UDFs were missing on FE followers after

RESTORE ... AS <new_db>

. #74313 • Queries could become unkillable when a coordinator held its lock during external resource cleanu… StarRocks/starrocks

Friendly reminder: Our webinar, Introducing StarRocks 4.1, is happening tomorrow at 10:00 AM PT! yaySave your seat here: bit.ly/4w0J936 <!channel> Join us for a walkthrough of what's new in 4.1 and what it means for your cluster — then bring your questions for the live Q&A. We'll cover adaptive data distribution with automatic tablet split and merge (no more hand-tuning bucket counts), large-capacity tablets that cut FE metadata and scheduling overhead, Fast Schema Evolution v2 (now for materialized views too), deeper Iceberg integration, end-to-end cache observability, and SQL upgrades like recursive CTEs and skew join v2. Can't join live? Register anyway and we'll send the recording afterward!

Hi community, 👋 Here are community reading recommendations for building faster, more efficient analytics pipelines with StarRocks Real-time analytics and lakehouse workloads often come with their own challenges — from improving query performance on Iceberg tables to building reliable streaming data pipelines. Here are two practical guides from the StarRocks community: 🚀 Replacing Presto with StarRocks for Faster Ad-Hoc Analytics on Iceberg Learn how one EdTech platform reduced CPU usage by 77% and improved P90 query latency by 2–3× by migrating their Iceberg analytics workloads to StarRocks. Read more: medium.com/starrocks-engineering/replacing-presto-with-starrocks-for-faster-ad-hoc-analytics-on-iceberg… ⚡ Designing a Practical Kafka to StarRocks Pipeline for Near Real-Time OLAP Explore how to build a Kafka → StarRocks pipeline to power near real-time analytics workloads. Read more: medium.com/starrocks-engineering/designing-a-practical-kafka-to-starrocks-pipeline-for-near-real-time-olap… Have a use case or architecture you’d like to discuss? Comment in the threads👇

🏅 Introducing the StarRocks Community Badges Ver.2026! We’ve launched a new badge program to recognize the amazing people who help build and grow the StarRocks community. There are two badge paths: 🛠️ Contributor – for technical contributions such as merged PRs, documentation improvements, bug reports, testing, and ecosystem projects. 🌟 Champion – for community contributions such as blogs, talks, meetups, tutorials, demos, case studies, and helping other community members. Each path currently has two badge levels, plus one hidden level as the highest honor. The hidden badge will be unlocked at the end of the year to recognize individuals who have made exceptional contributions to the StarRocks community. If you’ve contributed to StarRocks, we’d love to recognize your work with a digital badge that you can add to your LinkedIn profile or share on social media. 👉 Apply here: forms.gle/… The badge program recognizes contributions made during the 2026 calendar year. If you know someone in the community who’s made great contributions, please share this with them. Thank you for helping make StarRocks better! ❤️

Release - 4.0.13 New release published by wangsimo0 4.0.13 Release Date: July 16, 2026 Behavior Changes • The escape handling of

LIKE

predicates with constant operands (folded on the FE) now matches MySQL 8:

SELECT 'a\\b' LIKE 'a\\\\b'

returns

1

and

SELECT 'a\\b' LIKE 'a\\b'

returns

0

. Queries that relied on the previous non-MySQL escaping semantics will return different results. #74814 •

SHOW [FULL] FUNCTIONS

now always includes the

isolation

property (

shared

or

isolated

) in the Properties column of UDFs, so users can tell whether the property is set without recreating the function. #75255 • Iceberg REST catalogs with vended credentials use the table metadata cache again, reverting the earlier cache bypass that sent every

getTable()

to the REST catalog and caused AWS Lake Formation

Rate exceeded

failures. Cached tables now renew their credentials on every refresh cycle, and the table cache expiry for REST catalogs is additionally capped at 3000 seconds. #75431 Improvements • Added checksum protection for shared-data tablet metadata and transaction logs. #74924 • Supported combined transaction log / file bundling for

FRONTEND_STREAMING

loads. #74460 • Scoped shared-data schema-change job locks to the table to reduce lock contention. #75087 • Batch tablet force-delete marking now acquires the

TabletInvertedIndex

write lock once per batch instead of once per tablet. #75616 • Added an FE metric for the maximum pending-publish time of committed transactions. #75025 • Added a memory limit check for column upgrades in window operator processing. #75821 • Removed unnecessary per-row seeks in the offsets-only read path of array columns. #75861 • Foreground row-count estimation of Iceberg tables is now computed from manifest metadata without enumerating every data file. #75280 • Addressed security vulnerabilities (CVE): excluded the vulnerable

org.jline:jline

(jline-remote-telnet) from Hadoop transitive dependencies, and upgraded jackson-databind to 2.21.4. #75066 #75373 Bug Fixes The following issues have been fixed: • Incremental scan-range scheduling could recompute a different per-driver layout when reusing a deployed fragment instance, leaving part of the scan ranges unconsumed and losing rows (for example, INSERT from Hive). #74674 • INT96 timestamps nested inside ARRAY/MAP/STRUCT in Parquet files read via

FILES()

or Broker Load missed the session-timezone conversion and were returned shifted by the timezone offset (top-level INT96 columns were correct). #74868 • The audit log did not record the exported row count of

SELECT INTO OUTFILE

. #74467 • A strict cast could raise an overflow error from the underlying data of NULL rows, which should be ignored. #74903 •

parse_json

did not respect the

ALLOW_THROW_EXCEPTION

setting when handling invalid input. #74976 • First-load statistics collection could not be enabled per table while disabled globally: an explicitly set table property now takes precedence over the global configuration

enable_statistic_collect_on_first_load

. #74794 • Partial column updates on shared-data tables could crash the BE or silently corrupt data when the tablet schema drifted from the transaction schema. #74005 • Unexpected BE process restarts. #74424 • A BE crash (SIGFPE) in Iceberg `truncate`/`bucket` partition transforms when the width or bucket count is zero. #74998 • A BE crash caused by a null

driver_executor

in

FragmentContext::set_final_status

. #75030 • A race between transaction begin and autovacuum could delete a still-needed transaction log, permanently wedging the partition's publish ("Both txn_log and corresponding tablet_meta missing"). #74906 •

avg(DISTINCT x)

was incorrectly rewritten to use a sum/count materialized view, returning wrong results. #75071 • A boundary bug in TopN RANK sorting could produce incorrect results. #75045 •

split

/

split_part

/

str_to_map

with an empty delimiter could read out of bounds on invalid UTF-8 input. #75068 •

ALTER TABLE ... MODIFY COLUMN ... AFTER

a nonexistent column now returns a clear error message. #75073 • A BE crash (SIGFPE) in

mod()

/

pmod()

when computing the type's minimum value modulo -1. #74980 •

bar()

grew memory without bound on a negative or huge width (potential DoS); such inputs are now rejected with an error. #75143 • The transaction-state callback was not unregistered when a multi-statement stream load task was removed. #75188 • A crash when a query was cancelled during spill partition sorting. #75140 • The query memory limit was not enforced during table function execution. #75179 • Selecting a column named

floor

or

ceil

failed at parse time with a ClassCastException. #75241 • A heap-use-after-free in

OrderedPartitionExchanger

when the previous chunk was mutated downstream. #75279 • Three FE metadata-lock correctness races. #74968 • Load spilling could dereference a missing query context when recording spill metrics. #75236 • ADLS2

ListPaths

on storage accounts without hierarchical namespace caused CN crashes and vacuum failures. #75166 • BE/CN JVM metrics emitted invalid Prometheus

# TYPE

lines. #75240 • JIT code generation truncated LARGEINT literals to 64 bits, producing wrong results. #75137 • A combined ALTER TABLE on an external Iceberg table re-executed already-queued actions. #74036 • A nested-loop join crash caused by a build-side column nullability mismatch. #75343 • Iceberg manifest column statistics are now cached selectively to avoid excessive FE memory usage. #75395 •

addPhysicalPartition

could create only one physical partition per call, making physical-partition backfill of random-distribution tables extremely slow. #75430 • A SQL injection vulnerability in the `informatio… StarRocks/starrocks

Release - 3.5.20 New release published by yingtingdong Release date: July 23, 2026 Behavior Changes •

CREATE DATABASE IF NOT EXISTS

on Iceberg REST catalogs now succeeds silently when the database already exists, instead of raising an error. #75017 • Iceberg REST catalogs with vended credentials now cache

Table

objects and keep their credentials refreshed on access, instead of bypassing the cache and re-fetching from the REST catalog/Lake Formation on every

getTable()

call, which could trigger AWS

Rate exceeded

errors. #75431 • GIN inverted-index-accelerated

NOT MATCH

predicates no longer return rows with a

NULL

value, matching SQL three-valued-logic semantics. #75578 Improvements • Added the FE metric

txn_max_committed_pending_publish_ms

, a per-database gauge reporting the longest time a committed transaction has been pending publish, to help diagnose stuck or lagging version publishing. #75025 • Enforced the query memory limit when a column is upgraded (widened) during window-function aggregation in

Analytor

, instead of letting it grow unbounded. #75821 • Removed useless per-rowid seeks in the array-column offsets-only read path used by `array_length()`/`cardinality()`. #75861 Bug fixes The following issues have been fixed: • Several wrong-result issues:

EliminateSortColumnWithEqualityPredicateRule

dropping the global

LIMIT

under concurrency;

SplitJoinORToUnionRule

producing duplicate rows for a null-safe-equal (

<=>

) `JOIN ON p1 OR p2`; JIT codegen truncating

LARGEINT

literals

>= 2^64

to 64 bits; `array_map`/`transform` silently dropping

NULL

rows when all non-null input arrays were empty; nested dictionary expressions rebuilt inconsistently across exchange fragments causing dict-decode failures; and a

LIKE

pattern with the

_

wildcard returning wrong rows on a GIN inverted index. #74983 #75038 #75137 #75141 #75246 #75551 • Join-reorder column pruning could drop a column still referenced by a predicate, causing a

missing statistic of col

planning error, and

JoinTuningGuide

could lose

predicateCommonOperators

when rebuilding a join, failing plan validation. #74791 #75773 • Sync materialized view/rollup rewrite could lose a rollup column when a query aggregated the same base column twice (e.g.

min(c)

and

max(c)

), and async materialized view rewrite could serve stale results after an Iceberg base table's

rollback_to_snapshot

. #75528 #75924 •

PARTITION-TOP-N

could rewrite its partition-by column to a dictionary slot that no longer existed, failing with a

slot_id not found

error. #75956 • An NPE collecting view tables when a

SECURITY INVOKER

view's stored definition contains a CTE. #74813 • Three FE metadata-lock correctness races around

DROP PERSISTENT INDEX

,

RestoreJob

post-restore handling, and related unlocked paths. #74968 • A race between FE EOS-cancel and BE stage-2 deploy could mark a fully successful query as canceled. #75009 •

ApplyTuningGuideRule

could throw

UnsupportedOperationException

when an earlier rewrite produced an

OptExpression

with an immutable input list. #70785 • BE/CN crashes: a null

driver_executor

when a cancel RPC arrives before pipeline start; a use-after-free in the spill partition-sort-sink cancel path; a heap-use-after-free in

OrderedPartitionExchanger

for a skew-hinted window function at DOP>1; an

NLJoin

crash from a build-side column nullability mismatch; a

StructColumn

field-count mismatch in

UNNEST

output; a crash loop reading a flat-JSON column that changed from

NOT NULL

to nullable during compaction; an uncaught memory-allocation exception in `NLJoinProbeOperator`; a crash in primary-key auto-increment partial-update apply; and a crash rewriting predicates inside an

array_map

lambda during scan-predicate pushdown. #75030 #75140 #75279 #75343 #75445 #75680 #75788 #76119 #76380 •

histogram()

crashed (or silently mis-bucketed values) on a non-positive

bucket_num

instead of raising a clear error, and

bar()

could grow an unbounded string for a negative or huge

width

argument, exhausting BE memory. #75041 #75143 • A query using

unnest

over array columns could exceed

query_mem_limit

and get the BE OOM-killed instead of failing just that query. #75179 • A second-order SQL injection in the

information_schema.task_runs

`TASK_NAME`/`QUERY_ID` predicate lookup. #75520 •

SHOW CREATE ROUTINE LOAD

could emit a spurious leading comma before the first load-desc clause, and an unescaped

jsonpaths

value, producing non-runnable DDL. #75522 #75755 • Shared-data (lake)

SHOW PARTITIONS

and

information_schema.partitions_meta

reported every physical partition's bucket count as the table-level default instead of its own bucket count. #75734 • Several dependency CVEs by upgrading

jackson-databind

and Netty. #75373 #76555 • Batched

TabletInvertedIndex

write-lock acquisition in

markTabletsForceDelete

, reducing lock churn when force-deleting many tablets at once. #75616 • Batched tablet inverted-index writes in the insert-overwrite path. #75923 • Skipped an unnecessary remote

clear_parent_path

call when a load spill never used remote storage. #76224 • A null-padding size mismatch for missing columns in

ParquetScanner

so padded rows match the actual per-batch chunk size instead of the whole Parquet/Arrow batch size. #75981 • Vulnerable, stale transitive dependencies (old BouncyCastle, OkHttp 2.x, Tomcat, and others) that previously shipped alongside their fixed counterparts. #76097 StarRocks/starrocks

Friendly reminder: Our webinar, Real-Time Analytics Has an Update Problem, is happening tomorrow at 10:00 AM PT! yay Save your seat here: bit.ly/4hFi54R <!channel> Join us to see what an update actually costs on columnar storage, why merge-on-read slows every query as updates pile up, and the architecture that keeps data freshness under 5 seconds without impacting query performance — then bring your questions for the live Q&A. Can't join live? Register anyway and we'll send the recording afterward!

[🆕 Blog Post] 🔎 How do you build hybrid search with StarRocks? Real-world AI retrieval rarely relies on vector similarity alone. It often needs to combine semantic search with full-text matching, structured filters, and business-specific ranking. Our latest technical deep dive explores how StarRocks brings these paths together in one Search + OLAP engine. It covers: • Running vector, full-text, and scalar retrieval in a unified query path • Choosing between pre-filtering and post-filtering • Fusing results with RRF, weighted scoring, or custom rerankers • Supporting multimodal lakehouse search with StarRocks and Apache Paimon If you’re building RAG, multimodal search, or AI data-retrieval pipelines, take a look: 👉 Read the blog: medium.com/towards-data-engineering/from-data-lake-to-multimodal-lakehouse-building-hybrid-retrieval-for-ai… What hybrid-search use cases are you working on with StarRocks? We’d love to hear about them in the thread. 👇

Hi Community, 👋 Are you currently using Postgres? 📊 PostgreSQL is great for transactional workloads—but as data volumes and analytical queries grow, performance can hit a wall. This practical guide shows how to build a real-time CDC pipeline from PostgreSQL to StarRocks using Debezium and Kafka, covering setup, data consistency, and performance testing. 👉 Read the step-by-step guide

Friendly reminder: Our webinar, Unpredictable Queries, Predictable Latency: Architecting Joins for Analytical Agents, is happening tomorrow at 10:00 AM PT! danceml Save your seat here: bit.ly/4wyR2vT <!channel> Join us to see what an analytical agent actually sends to a database, why precomputation can't cover it, and which architectural pieces produce a low-latency join when nothing has been prepared in advance — then bring your questions for the live Q&A! Can't join live? Register anyway and we'll send the recording afterward!

Hey everyone! 👋 We’re co-hosting a meetup with TRM Labs in San Francisco🌉 on Sept. 3: Scaling the Data Lake for the AI Era. Rahul Raina, CTO from TRM Labs and Sida Shen, Product Manager from PhoenixAI will share real-world lessons on using StarRocks for petabyte-scale analytics, including architecture, performance, and query optimization. What you can expect: • Real-world lessons from operating analytics at scale • Practical discussion on architecture and performance • Open Q&A with the speakers • Time to connect with data, ML, and infrastructure engineers across the Bay Area If you’re in the Bay Area, we’d love to see you there! 📅 Sept. 3 🕔 500–730 PM 📍 SHACK15, San Francisco 👉 Register here: luma.com/trmlabs-3kox

Release - 4.0.14 New release published by wangsimo0 4.0.14 Release Date: August 11, 2026 Behavior Changes • A static partition clause in an

INSERT

into an Iceberg or Hive table is now validated even when a target column list is given. A clause that names a column which is not a partition column of the table is rejected with a clear error instead of being silently ignored. Every

INSERT

shape that was accepted before is still accepted. #76659 •

ARRAY

and

MAP

constructors now fail with a

CapacityLimitExceed

error when the flattened result of a single chunk exceeds the addressable byte limit, instead of silently returning corrupted values that

CREATE TABLE AS SELECT

,

INSERT

, or a materialized view refresh could persist. For affected queries, reduce

chunk_size

or split the constructor expression. #76419 • Division expressions whose divisor is not constant (for example,

10 DIV c

) are no longer treated as monotonic, which stops unsafe Zone Map pruning from dropping rows that satisfy the predicate. Queries filtering on such an expression may return more rows than in earlier versions; the previous results were incorrect. Monotonic forms such as

c DIV 10

still benefit from the pruning. #76744 • On a column with a GIN inverted index,

NOT MATCH

no longer returns rows whose value is

NULL

, which brings it in line with SQL three-valued logic. Queries that depended on the previous behavior return fewer rows. #75578 • Under

"compression" = "zstd"

, flat JSON sub-columns and the synthetic null and offset sub-columns of

ARRAY

,

MAP

, and

STRUCT

columns are now compressed. They were previously written as raw pages, which could make a ZSTD table larger on disk than the same table compressed with LZ4. Only segments written after the upgrade are affected, so existing tables shrink gradually as their segments are rewritten. #76949 • The Iceberg partition cache is now bounded by memory instead of by entry count, through the new catalog property

iceberg_partition_cache_memory_usage_ratio

(default

0.1

), and its footprint is reported in

/api/memory_usage

and the minute-level memory logs. Under memory pressure the cache may hold fewer entries than before; raise the ratio to restore the previous footprint. #76165 • FE memory estimation now also counts the internal overhead of maps and collections — per-entry nodes and the backing table — instead of only the sampled keys, values, and elements. Weight-bounded caches such as the Iceberg metadata caches therefore report their true footprint and evict earlier for the same

*_memory_usage_ratio

, which lowers FE memory usage but may increase cache misses. #75971 • When the keep-alive GC reaps an expired external scan context — left behind by a Spark or Flink connector reader that died without calling

close_scanner

— it now cancels the corresponding pipeline fragment. The buffered scan memory is released within

keep_alive_min

plus one GC interval instead of staying pinned until

query_timeout

. #76535 • The audit log of a statement forwarded to the Leader FE now records the relations resolved by the Leader, so

QueriedRelations

holds fully qualified table names with CTE references excluded, matching what the Leader logs. A Follower falls back to local collection only when the statement ran locally or the Leader did not return the list, for example during a rolling upgrade. #76387 • The

Operation

column of

SHOW ALTER TABLE OPTIMIZE

now shows a readable description of the optimize operation instead of an internal object address such as

com.starrocks.sql.ast.OptimizeClause@b5dc069

. #75948 Improvements • Supported complex types (

ARRAY

,

MAP

, and

STRUCT

) in Paimon tables; querying such columns could previously crash the BE. #66784 • Added the

LAST_REFRESH_TIME

column to

information_schema.materialized_views

and

SHOW MATERIALIZED VIEWS

. It reports the data-freshness timestamp used by the

mv_rewrite_staleness_second

check, which is distinct from the existing

LAST_REFRESH_FINISHED_TIME

(when the refresh job finished running). #71642 • Added the mutable BE configuration item

object_storage_client_cache_size

(default

8

), which replaces the hard-coded capacity of the S3 and Azure Blob client caches. #75851 • Metadata refresh of a filesystem-backed external table triggered by

INSERT ... SELECT

no longer runs while FE internal metadata locks are held, so slow remote metadata access no longer stalls unrelated work on the same path. #73391 • Multi-statement (multi-table) transaction Stream Load now dispatches all per-table channels before waiting for them, instead of dispatching and waiting for each table in turn, which shortens commit time for CDC pipelines that write many tables under one label. #76715 • Reduced lock contention across the

INSERT OVERWRITE

code path by relaxing or shortening the table-level locks held in each phase. #75828 • Error messages for the large-column capacity limit no longer embed internal diagnostics such as the driver address and the operator chain, and the

Capaticy

typo in the shared status string is corrected. #76303 • Shared-data Primary Key publish now applies the per-segment delete vectors carried in

op_write.seg_delvecs

, so a BE running this version consumes such metadata correctly instead of leaving duplicate primary-key rows live. #76474 • Addressed security vulnerabilities (CVE): upgraded Thrift to 0.24.0 and Netty to 4.1.136.Final, bumped the PostgreSQL JDBC driver to 42.7.12, and removed vulnerable transitive dependencies that shipped alongside their fixed counterparts, namely

bcprov-jdk15on

, the end-of-life OkHttp 2.x line,

avro-ipc

(which bundles jQuery 1.4.2), and the Jetty client and security jars. #76922 #76555 #76783 #76097 #76270 Bug Fixes The following issues have been fixed: • The query cache could store an incomplete per-tablet result and serve it to later queries, returning wrong results. #77066 #77404 •

COUNT(DISTINCT)

on an Iceberg table partitioned by

bucket()

returned an over-count when

enable_bucket_aware_execution_on_lake

was enabled and the

GROUP BY

list was a superset of the bucket column. #76601 • JSON subfield pushdown returned wrong results when two subfield keys of the same column differed only in case (for example,

get_json_string(c, 'Campaign')

and

get_json_string(c, 'campaign')

), because the generated column names are resolved case-insensitively. Such collisions are now excluded from the pushdown. #76594 #76593 •

array_difference

on integer input computed each adjacent difference in the 32-bit input typ… StarRocks/starrocks

Release - 3.5.21 New release published by yingtingdong Release date: August 28, 2026 Behavior Changes • Reverted the 3.5.20 change that cached Iceberg REST vended-credential tables and kept their credentials fresh: on

branch-3.5

, that caching caused

INSERT OVERWRITE

transactions against a slow Iceberg REST catalog to stay

COMMITTED

but not

VISIBLE

for several seconds, because a REST metadata refresh could now run while the planner held a lock that transaction publishing also needs. The caching behavior remains in place on the 4.x line. #77039 • GIN (inverted) indexes on Primary Key tables now read from the correct segment after a column-mode partial update, instead of serving stale index data from the unmodified base segment. #76271 • Audit logs for statements forwarded to the leader FE now record the leader-resolved, fully qualified table relations (with CTEs excluded), instead of the follower's unresolved names. #76387 • `ARRAY`/`MAP` constructor expressions now reject a result whose flattened size exceeds 4 GB instead of silently wrapping around and returning corrupted values. #76419 • Iceberg REST catalogs using OAuth2 client-credentials now self-heal their background token-refresh session after a prolonged failure, instead of leaving the catalog permanently unable to refresh its access token. #76457 • External scan contexts (for example, an abandoned Spark/Flink connector read) now properly cancel their pipeline fragments when reaped as expired, instead of leaving them running. #76535 • External scan plans (Spark/Flink connector reads) now set

query_delivery_timeout

, so their

QueryContext

no longer waits indefinitely for fragments that will never arrive. #76536 •

array_difference()

on integer arrays no longer overflows in 32-bit precision before widening to

BIGINT

, fixing incorrect results when the true difference falls outside the

INT

range. #76569 • Division expressions with a non-constant divisor (for example,

10 DIV c

) are no longer treated as monotonic, fixing incorrect ZoneMap-based pruning that could return an empty or wrong result. #76744 • The configured compression codec now applies to the synthetic null/offset sub-columns of flat JSON,

ARRAY

,

MAP

, and

STRUCT

columns, instead of always writing them as uncompressed raw pages. #76949 • Load quorum selection no longer picks a replica in

DECOMMISSION

state as the load primary. #77035 • Optimizer rules that rebuild a logical window operator now preserve its

inputIsBinary

flag, keeping the binary-input merge optimization for ranking-window pre-aggregation intact. #77058 Improvements • Materialized views no longer force a full-partition refresh when manually set from

INACTIVE

to `ACTIVE`; only the metadata version map is cleared during a schema change. #57371 • Improved error messages for large-column-capacity-limit checks by removing internal diagnostics (raw pointers, operator dumps) from the user-facing error, and fixed a typo in the shared status string. #76303 Bug fixes The following issues have been fixed: • Guarded against cyclic view definitions: an

ALTER VIEW

that closes a reference cycle (

v1

->

v2

->

v1

) is now rejected with a clear error instead of causing a later

SELECT

to recurse forever and crash with

StackOverflowError

. #75033 • Forbade pushing an aggregate down through a CASE expression that has a non-null constant ELSE branch, fixing a planning abort (

IllegalStateException

) that could occur once the rule fired. #75037 •

CTEAnchor

now prunes correctly when its child is a

ValueOperator

. #64491 • Fixed predicate conversion for Paimon: an AND-combined predicate now retains its convertible not-null branch instead of the whole conversion returning null. #66038 • The Arrow Flight prepared-statement schema no longer reports every view column as nullable regardless of its actual

NOT NULL

definition, and a related regression that produced wrong nullability for `GROUP BY ROLLUP`/`CUBE`/`GROUPING SETS` key columns has also been fixed. #75684 #76149 • Fixed two statistics bugs: an OR-predicate statistics estimate that always clamped the merged

nullsFraction

to

1

regardless of the real value, and a column-statistics cache load failure under

ERROR_IF_OVERFLOW

when a column's persisted min/max is an empty string. #75864 #76684 • Fixed a bRPC stub cache timer leak that leaked memory over time. #75973 •

UNNEST

output struct pruning now uses its input array's subfield group instead of the output's own, fixing a mismatch between the struct type BE materializes and the one FE declares. #76002 • Fixed Arrow Flight Prepared Statement forwarding sending the wrong action-type string when a request is forwarded to a different FE, which caused every ADBC client using prepared statements behind a load balancer to fail. #76310 • Hive

getTable()

now reconnects before falling back to

get_table_req()

, fixing an intermittent

out of sequence response

/

Unknown table

error when querying an Iceberg table through a Hive metastore catalog. #76456 • Made catalog-drop existence checks atomic under the write lock, fixing a check-then-act race that could persist a redundant drop record when two drops of the same catalog ran concurrently. #76778 • Fixed

PipeObservable

emitting a source event instead of a sink event on a deferred sink notification, which could leave a driver blocked on

OUTPUT_FULL

unresponsive. #76782 •

dictionary_get()

no longer rejects a non-NULL key when its input column's cached

has_null

flag is stale. #76881 • Added the missing

arrow-compression

module for Arrow Flight SQL, restoring LZ4/ZSTD codec support for compressed Arrow IPC clients. #76921 • Fixed an OOM in streaming pre-aggregation under memory pressure with spill enabled. #76702 •

Set

operators are no longer placed in colocate execution groups, fixing a hang caused by their branches being terminated by a plain local-exchange sink instead of a grouped-execution sink. #77025 • Stopped the query cache from storing incomplete per-tablet results for an aggregation with a

LIMIT

, which could return wrong results from the cache. #77066 • An insert-overwrite failure is no longer journaled against a table that was concurrently dropped, fixing an FE crash on journal replay. <https://github.com/StarRocks/starrocks/pull/77… StarRocks/starrocks

📚 New Blog: Building a Streaming Lakehouse at Alibaba with Fluss, Paimon, and StarRocks How do you query seconds-fresh streaming data, minute-level lakehouse data, and years of history through one analytical interface? In this production deep dive, Zhu Ao, Senior Data Engineer at Taotian Group, explains how Alibaba combines: • Fluss for seconds-fresh streaming data • Apache Paimon for minute-level and historical storage • StarRocks as the unified, high-performance OLAP query layer The results include an 80%+ reduction in consumer bandwidth and deserialization costs, a 50%+ improvement in development and operational efficiency, and a development and validation cycle shortened from five days to two. 👉 Read the complete architecture breakdown: medium.com/starrocks-engineering/building-a-streaming-lakehouse-at-alibaba-with-fluss-paimon-and-starrocks…

Friendly reminder: Our webinar, Apache Iceberg at Sub-Second Speed: Architecting the Serving Layer, is happening today at 10:00 AM PT! Save your seat here: bit.ly/4qPCMhf <!channel> Eric Sun will be joining Sida Shen live, bringing firsthand perspective from his experience leading Data Platform + Datastores at Coinbase. Together, they’ll explore how to serve sub-second analytics directly on Apache Iceberg, where performance and serving challenges emerge in production, and what it takes to make it work at scale — without bringing back separate serving copies and sync pipelines. Bring your questions for the live Q&A! Can’t join live? Register anyway and we’ll send the recording afterward. cat dance

📣 The Call for Papers for NEXT QUERY 2026 is now open! <!channel> The way we interact with data is changing. What will the next query look like? Who—or what—will ask it? And how will the answer turn into action? That’s the idea behind NEXT QUERY 2026, the Data & AI Infrastructure Summit presented by PhoenixAI. On November 4, data practitioners, engineers, and developers will come together to share how they’re building analytics systems for today—and preparing them for what comes next. This year’s theme is Building the Context for What’s Next. As AI agents enter analytical workflows, better models alone aren’t enough. Every query also needs the right business meaning, knowledge, memory, and governance to produce answers people can trust. We’re expanding beyond a traditional StarRocks community summit and inviting the broader Data & AI community to share practical stories about: • Real-time analytics and data applications • Lakehouse architectures • Performance, scalability, and governance • AI and agentic analytics use cases Whether you’re running StarRocks in production or building the next generation of data and AI systems, we’d love to hear what you’ve learned. 🗓️ Submission deadline: September 18 👉 Learn more and submit your proposal: docs.google.com/forms/…/viewform

Hi everyone! 👋 Before an AI agent queries your Iceberg lakehouse, it needs to know which data and definitions to trust. Explore the roles of Polaris, OpenMetadata, and StarRocks in building trusted context for AI. 📖 Read the article Join us at NEXT QUERY 2026, the Data & AI Infrastructure Summit presented by PhoenixAI, on Nov 4, 2026. We’ll explore how metadata, semantics, lakehouse architecture, and agentic analytics come together to build the context behind every query. Registration is now open! 🎟️ Register for NEXT QUERY

Friendly reminder that our webinar, PhoenixAI Anywhere: Sub-Second Analytics on Live Data, Inside Your Own Network, is happening today at 10 AM PT / 1 PM ET. <!here> We’ll show how PhoenixAI runs complex, multi-table SQL on normalized data at sub-second latency, with streaming updates queryable within seconds, and how PhoenixAI Anywhere brings that performance into your own Kubernetes environment, including private and air-gapped deployments with no outbound dependency. The session will also include a walkthrough of the Anywhere Console used to deploy and manage it. If you’ve been curious about what PhoenixAI adds on top of StarRocks or what a self-managed PhoenixAI deployment looks like in practice, this is a good chance to see it live. danceml Save your seat: bit.ly/3TelPAq Can’t join live? Register anyway and we’ll send you the recording afterward.

Release - 4.1.3 New release published by kc-phxai 4.1.3<#413|​> Release Date: July 14, 2026 Behavior Changes<#behavior-changes|​> •

CTAS

now preserves explicitly declared

VARCHAR(N)

column lengths instead of widening them to

VARCHAR(MAX)

. Existing tables are unaffected; new tables created with

CTAS

will enforce the declared length on subsequent writes. #73498 • Querying

sys.fe_memory_usage

or

sys.fe_locks

without the

OPERATE ON SYSTEM

privilege now returns a clear access-denied error instead of a misleading node-lookup failure. #73567 •

FILES()

and broker/stream load no longer apply a session timezone shift to

INT64

Parquet timestamps written with `isAdjustedToUTC=false`; those timestamps are now treated as wall-clock values and loaded as-is. Data loaded from such files before v4.1.3 may differ from data loaded after upgrading; reload if consistency is required. #73674 • Successfully committed multi-table transaction stream load jobs are now correctly shown as

VISIBLE

in

information_schema.loads

and

SHOW STREAM LOAD

instead of remaining stuck at

PREPARING

. #74386 • Connector incremental scan range scheduling now consistently reuses the deployed fragment's driver layout, preventing scan ranges from being incorrectly assigned to non-existent drivers. #74674 •

LIKE

constant-folding now matches MySQL 8 backslash-escape semantics, correcting cases where patterns such as

'a\\\\b'

previously returned opposite results. #74814 • Routine Load now supports the

property.kafka_partition_discovery

property, which allows partition auto-discovery to continue even when

kafka_partitions

and

kafka_offsets

are specified to seed exact starting offsets. When

property.kafka_default_offsets

is not set, the default starting offset for partitions discovered after the job already has consuming progress changes from

OFFSET_END

to

OFFSET_BEGINNING

— and this applies to all auto-discovery jobs, not only those using the new property. #74729 • Non-group-by aggregates are now pushed down through

UNION ALL

branches before merging, reducing network transfer and memory usage for queries that aggregate over a union. #73930 • IVM maintenance queries are now re-derived from the current view definition at each refresh instead of using the frozen query text stored at

CREATE

time; existing MVs automatically benefit from rewriter bug fixes without needing to be recreated. #74881 • Sample-based tablet pre-split now spreads pre-split shards across all compute nodes (

SPREAD

placement) instead of packing them onto the source tablet's worker (

PACK

placement), improving load parallelism. #75514 •

ALTER TABLE ... MODIFY COLUMN

that changes only the column comment now takes the lightweight metadata-only path instead of spawning a full schema-change job, and this now works on Primary Key columns as well. #75325 •

FLOOR

and

CEIL

are now treated as non-reserved keywords and can be used as column names without quoting. #75241 •

SHOW FUNCTIONS

output now always includes the

isolation

property (

shared

or

isolated

) in the Properties column for UDFs and UDAFs. #75255 • The default value of

lake_vacuum_min_batch_delete_size

is raised from 100 to 200, improving vacuum throughput on object storage by batching more stale-file deletions per

DeleteObjects

request. #74304 • Iceberg REST catalog tables with vended credentials are now cached and their credentials are refreshed in the background, eliminating per-

getTable()

GetDataAccess

calls that caused AWS Lake Formation rate limiting. #75431 • IVM

bitmap_union

,

hll_union

, and

percentile_union

aggregate states are now stored once in the materialized view instead of twice (visible column + hidden

__AGG_STATE_

column), halving storage for those sketch types. #75760 • Incremental materialized views now support

bitmap_agg

,

hll_union

,

percentile_union

, and

bitmap_union

aggregate functions, enabling exact distinct-count and sketch-based aggregations to be maintained incrementally. #75587 #75610 • Sample-based tablet pre-split tablet count is now rounded up to the nearest multiple of the active compute-node count for even distribution, and bounded below by a minimum tablet size to avoid excessive fragmentation on small loads. #75360 #75584 Improvements<#improvements|​> • The

ngram_search

function now accepts a non-constant needle argument. #74675 • Added an HTTP authentication framework controlled by the

enable_http_auth

FE configuration, gating authentication and RBAC enforcement on all external HTTP endpoints. #73822 • Added refresh and placement observability columns (

refresh_warehouse

,

refresh_resource_group

,

refresh_mode

,

refresh_type

,

last_refresh_details

) to

information_schema.materialized_views

. #74342 • Added opt-in lazy refresh of the external statistics cache on journal replay, controlled by a new FE config, to prevent a slow or stuck external metastore from stalling FE journal replay or startup. #74371 •

VARCHAR

length increase is now allowed on range-distribution (shared-data) sort-key columns via fast schema evolution without a data rewrite. #74698 • Added a stack-trace dump when a shared-data transaction log write exceeds a configurable threshold, making slow

put_txn_log

/

put_combined_txn_log

calls easier to diagnose. #74704 • Tablet pre-split meta-tier footer readers now support

DATE

,

DATETIME

,

DECIMAL

,

VARCHAR

, and ORC

TIMESTAMP

sort keys, reducing the number of loads that must fall back to data-tier sampling. #74710 #74739 #74792 #74902 #74955 #75186 #75209 #75427 #75697 • Sample-based tablet pre-split now applies to

INSERT INTO ... SELECT ... FROM <OLAP table>

loads, and also to column-list

INSERT

statements that include all sort-key columns. #74828 #75345 • Added Adler-32 checksum protection for shared-data tablet metadata and transaction log files, enabling silent corruption to be detected on read. #74924 • Added the

txn_max_committed_pending_publish_ms

FE metric per database, reporting the age of the… StarRocks/starrocks