+1
Test environment: - OS: macOS (Darwin 25.2.0, ARM64/Apple Silicon) - Java: OpenJDK 21.0.1 LTS (TencentKonaJDK) - Fluss: 0.9.0-incubating (binary distribution) - Flink: 2.2.0 - Connector: `fluss-flink-2.2-0.9.0-incubating.jar` (built from 0.9 source) - Tiering JAR: `fluss-flink-tiering-0.9.0-incubating.jar` - Fluss Cluster: local-cluster mode (1 CoordinatorServer + 1 TabletServer) - Flink Cluster: Standalone (1 JobManager + 1 TaskManager, process=4096m) Tested Flink + Fluss Lakehouse Tiering (Paimon + Iceberg) with 40 test cases, all passed:*OK* - Tiering to Paimon — PK table (create/insert/tiering/$lake-query/$lake-agg/union-read/upsert/delete): *OK* (10 cases) - Tiering to Paimon — Log table (create/insert/tiering/$lake-query/$lake-agg/streaming-query): *OK* (5 cases) - Tiering to Paimon — Complex types PK table with ARRAY/MAP/ROW (create/insert/tiering/$lake-query/$lake-agg): *OK* (5 cases) - Tiering to Paimon — System table $lake$snapshots: *OK* (1 case) - Tiering to Iceberg — PK table (create/insert/tiering/$lake-query/$lake-agg/union-read/upsert): *OK* (8 cases) - Tiering to Iceberg — Log table (create/insert/tiering/$lake-query/$lake-agg/streaming-query): *OK* (5 cases) - Tiering to Iceberg — Complex types PK table with ARRAY/MAP/ROW (create/insert/tiering/$lake-query/$lake-agg): *OK* (5 cases) - Tiering to Iceberg — System table $lake$snapshots: *OK* (1 case) Detailed test cases: | # | Lake | Category | Test Case | Mode | Result | |---|------|----------|-----------|------|--------| | 1 | Paimon | PK Table | Create PK table with datalake enabled (freshness=10s) | batch | *OK* | | 2 | Paimon | PK Table | Insert 5 rows into PK table | batch | *OK* | | 3 | Paimon | PK Table | Point query (order_id=1) | batch | *OK* | | 4 | Paimon | PK Table | $lake query (read from Paimon storage) | batch | *OK* | | 5 | Paimon | PK Table | $lake aggregation (COUNT + SUM) | batch | *OK* | | 6 | Paimon | PK Table | Union Read (Fluss + Paimon) | batch | *OK* | | 7 | Paimon | PK Table | Upsert update (id=1) | batch | *OK* | | 8 | Paimon | PK Table | Verify upsert (point query id=1) | batch | *OK* | | 9 | Paimon | PK Table | Delete record (id=5) | batch | *OK* | | 10 | Paimon | PK Table | Verify delete ($lake COUNT) | batch | *OK* | | 11 | Paimon | Log Table | Create Log table with datalake enabled | batch | *OK* | | 12 | Paimon | Log Table | Insert 4 rows into Log table | batch | *OK* | | 13 | Paimon | Log Table | $lake query Log table | batch | *OK* | | 14 | Paimon | Log Table | $lake aggregation (GROUP BY event_type) | batch | *OK* | | 15 | Paimon | Log Table | Streaming query Log table (earliest) | streaming | *OK* | | 16 | Paimon | Complex Types | Create PK table with ARRAY/MAP/ROW + datalake | batch | *OK* | | 17 | Paimon | Complex Types | Insert complex type data (3 rows) | batch | *OK* | | 18 | Paimon | Complex Types | Point query complex types (id=1) | batch | *OK* | | 19 | Paimon | Complex Types | $lake query complex types | batch | *OK* | | 20 | Paimon | Complex Types | $lake aggregation complex types (COUNT) | batch | *OK* | | 21 | Paimon | System Table | $lake$snapshots (snapshot_id, commit_user, total_record_count) | batch | *OK* | | 22 | Iceberg | PK Table | Create PK table with datalake enabled (freshness=10s) | batch | *OK* | | 23 | Iceberg | PK Table | Insert 5 rows into PK table | batch | *OK* | | 24 | Iceberg | PK Table | Point query (order_id=1) | batch | *OK* | | 25 | Iceberg | PK Table | $lake query (read from Iceberg storage, 5 rows with __bucket/__offset/__timestamp) | batch | *OK* | | 26 | Iceberg | PK Table | $lake aggregation (COUNT=5, SUM=808.39) | batch | *OK* | | 27 | Iceberg | PK Table | Union Read (Fluss + Iceberg) | batch | *OK* | | 28 | Iceberg | PK Table | Upsert update (id=1, customer→Alice_Updated) | batch | *OK* | | 29 | Iceberg | PK Table | Verify upsert (point query id=1, Alice_Updated confirmed) | batch | *OK* | | 30 | Iceberg | Log Table | Create Log table with datalake enabled | batch | *OK* | | 31 | Iceberg | Log Table | Insert 4 rows into Log table | batch | *OK* | | 32 | Iceberg | Log Table | $lake query Log table (4 rows with timestamps) | batch | *OK* | | 33 | Iceberg | Log Table | $lake aggregation (GROUP BY event_type: click=2, view=1, buy=1) | batch | *OK* | | 34 | Iceberg | Log Table | Streaming query Log table (earliest, 4 rows) | streaming | *OK* | | 35 | Iceberg | Complex Types | Create PK table with ARRAY/MAP/ROW + datalake | batch | *OK* | | 36 | Iceberg | Complex Types | Insert complex type data (3 rows) | batch | *OK* | | 37 | Iceberg | Complex Types | Point query complex types (id=1, tags=[vip,active]) | batch | *OK* | | 38 | Iceberg | Complex Types | $lake query complex types (3 rows with ARRAY/MAP/ROW) | batch | *OK* | | 39 | Iceberg | Complex Types | $lake aggregation complex types (COUNT=3) | batch | *OK* | | 40 | Iceberg | System Table | $lake$snapshots (snapshot_id, operation=append/overwrite, summary) | batch | *OK* | Notes: - Paimon: 1.3.1 (`paimon-flink-2.0-1.3.1.jar` + `paimon-bundle-1.3.1.jar`) - Iceberg: 1.10.1 (`iceberg-flink-runtime-2.0-1.10.1.jar`) - Iceberg required `hadoop-client-api-3.3.6.jar` + `hadoop-client-runtime-3.3.6.jar` (Hadoop 3.x) to resolve `FileSystem.openFile()` API compatibility - Iceberg required patching `LakeFlinkCatalog.java` to use 3-arg `createCatalog(String, Map, Configuration)` via reflection (Iceberg 1.10.1 API change) - Tiering Service ran as Flink streaming job, data verified via `$lake` virtual table queries reading directly from lake storage - All `$lake` queries returned correct data with Fluss metadata columns (__bucket, __offset, __timestamp) ForwardXu [email protected] 原始邮件 发件人：Yunhong Zheng <[email protected]> 发件时间：2026年2月13日 22:51 收件人：dev <[email protected]> 主题：Re: [VOTE] Release Fluss 0.9.0-incubating (RC1) +1 (binding) I have verified the following new introduced features: aggregate merge engine, auto increment column, add column, rebalance and the kv snapshot lease. All of these features passed in RC1. Looking forward to the final release of Fluss 0.9-incubating! Yours, Yunhong(Swuferhong) On 2026/02/13 13:10:22 Yang Wang wrote: > +1 (non-binding) > > - verified signatures and checksums: ok > - verified the source builds correctly: ok > - checked the LICENSE and NOTICE files are correct > - tested Rebalance feature with Flink 2.2.0 (1 coordinator + 3 > tablet-servers): ok > - interface validation (add_server_tag, rebalance, > list_rebalance, cancel_rebalance) > - effect verification (node offline migration, replica > distribution, leader distribution) > - data correctness across rebalance operations > - confirmed list_rebalance ClassCastException from rc0 is > fixed in rc1 > - tested Delta Join with Flink 2.2.0: ok > - verified DeltaJoin optimization works correctly for CDC > sources with table.delete.behavior='IGNORE' > - built and verified with official git tag > (v0.9.0-incubating-rc1) > - tested Paimon DV Union Read with Flink 1.20.3 + Paimon 1.3.1: > ok > - confirmed PR #2326 resolved rc0 issues where DV table Union > Read returned 0 rows or stale data > - Union Read correctly returns complete and up-to-date data > across multiple update scenarios > > Best, > Yang > > yuxia <[email protected]> 于2026年2月12日周四 20:03写道： > > > Hi everyone, > > > > Please review and vote on the release candidate #1 for the Apache Fluss > > version 0.9.0-incubating, as follows: > > [ ] +1, Approve the release > > [ ] -1, Do not approve the release (please provide specific comments) > > > > > > The complete staging area is available for your review, includes: > > > > The official source release and binary convenience releases to be deployed > > to: > > * > > https://dist.apache.org/repos/dist/dev/incubator/fluss/fluss-0.9.0-incubating-rc1 > > > > Helm charts are available on: > > * https://dist.apache.org/repos/dist/dev/incubator/fluss/helm-chart > > (NB: you have to build the Docker images locally with the version > > 0.9.0-incubating in order to test Helm charts) > > > > All the files are signed with the key with fingerprint 56A9F259A4C18F9C, > > you can find the KEYS file here: > > * https://dist.apache.org/repos/dist/release/incubator/fluss/KEYS > > > > All artifacts to be deployed to the Maven Central Repository: > > * https://repository.apache.org/content/repositories/orgapachefluss-1004/ > > > > Git tag for the release: > > * https://github.com/apache/fluss/releases/tag/v0.9.0-incubating-rc1 > > > > Git commit for the release: > > * > > > > https://github.com/apache/fluss/commit/d6fd1f1f607a2672bff5d18d5ca811bfa920bbd7 > > > > Website pull request for the release announcement blog post > > * https://github.com/apache/fluss/pull/2590 > > > > Upgrade note for the new release: > > * > > > > https://github.com/apache/fluss/blob/release-0.9/website/docs/maintenance/operations/upgrade-notes-0.9.md > > > > Docker images for the release candidate: > > * fluss: apache/fluss:0.9.0-incubating-rc1 > > > > Please download, verify and test. To learn more about how to verify: > > > > https://fluss.apache.org/community/how-to-release/verifying-a-fluss-release/ > > > > > > The vote will be open for at least 72 hours. It is adopted by majority > > approval, with at least 3 PPMC affirmative votes. > > > > > > Best regards, > > Yuxia > > >
