You are viewing a plain text version of this content. The canonical link for it is here.
Posted to commits@fluss.apache.org by ab...@apache.org on 2026/09/23 23:23:46 UTC
(fluss-blog) branch main updated: [docs] Address review feedback on the Fluss 1.0 release blog (#18)
This is an automated email from the ASF dual-hosted git repository.
fresh-borzoni pushed a commit to branch main
in repository https://gitbox.apache.org/repos/asf/fluss-blog.git
The following commit(s) were added to refs/heads/main by this push:
new 4ed21b0 [docs] Address review feedback on the Fluss 1.0 release blog (#18)
4ed21b0 is described below
commit 4ed21b0cac802d243b12dc812f52d9d3ebb4fda1
Author: Jark Wu <ja...@apache.org>
AuthorDate: Thu Sep 24 07:23:39 2026 +0800
[docs] Address review feedback on the Fluss 1.0 release blog (#18)
* [docs] Address review feedback on the Fluss 1.0 release blog
Clarify filter pushdown and replica placement, add batch mode to the
bitmap query, remove the immediate coordinator takeover claim, and
deduplicate contributor aliases while excluding dependabot.
Co-Authored-By: Codex <no...@openai.com>
AI-Model: gpt-6
AI-Contributed/Feature: 14/14
AI-Contributed/UT: 0/0
* remove ipolyzos
AI-Contributed/Feature: 0/2
AI-Contributed/UT: 0/0
---------
Co-authored-by: Codex <no...@openai.com>
---
blog/releases/1.0.md | 14 ++++++++------
1 file changed, 8 insertions(+), 6 deletions(-)
diff --git a/blog/releases/1.0.md b/blog/releases/1.0.md
index ee76009..d2d25ea 100644
--- a/blog/releases/1.0.md
+++ b/blog/releases/1.0.md
@@ -106,7 +106,7 @@ ALTER TABLE sensor_events SET (
For a query with `temperature > 30 AND location = 'warehouse-A'`, the TabletServer only needs to inspect a small amount of metadata, such as minimum values, maximum values, and null counts, to determine whether a batch could contain matching records. This requires no decoding or row-by-row scanning of the actual data. The additional server CPU cost is therefore negligible, while unnecessary network traffic and client-side processing can be substantially reduced.
-Filter pushdown works on Arrow-format log tables and data written after statistics are enabled. Pushdown works at batch granularity by skipping non-matching batches. Flink SQL applies the remaining row-level filter automatically while users of Java, Rust, Python, and C++ clients must apply it themselves.
+Filter pushdown works on Arrow-format log tables and only for data written after statistics are enabled. Flink SQL applies the remaining row-level filter automatically, while users of Java, Rust, Python, and C++ clients must apply it themselves.
Learn more: [Filter Pushdown](https://fluss.apache.org/docs/1.0/engine-flink/reads/#filter-pushdown).
@@ -256,7 +256,9 @@ JOIN user_dict
FOR SYSTEM_TIME AS OF e.proctime AS d
ON e.email = d.email;
--- 4. Aggregate across dimensions: merge bitmaps, then compute exact unique visitor counts
+-- 4. Aggregate across dimensions in batch mode: merge bitmaps, then compute exact unique visitor counts
+SET 'execution.runtime-mode' = 'batch';
+
SELECT channel,
rb_cardinality(rb_or_agg(uv_bitmap)) AS uv
FROM page_uv
@@ -337,7 +339,7 @@ Learn more: [Spark Reads](https://fluss.apache.org/docs/1.0/engine-spark/reads/)
### Coordinator High Availability
-Fluss 1.0 adds Coordinator high availability. When multiple Coordinators are deployed in the same cluster, one acts as the leader and the others remain on standby. If the leader fails, a standby immediately takes over metadata management and recovery coordination.
+Fluss 1.0 adds Coordinator high availability. When multiple Coordinators are deployed in the same cluster, one acts as the leader and the others remain on standby. If the leader fails, a standby is elected to take over metadata management and recovery coordination.
This reduces the management service's dependence on a single process. During rolling upgrades, Coordinators can be upgraded one at a time, keeping operations such as table management available while an individual Coordinator restarts.
@@ -345,7 +347,7 @@ Learn more: [Coordinator HA Deployment](https://fluss.apache.org/docs/1.0/instal
### Multiple Local Disks
-Fluss 1.0 supports configuring multiple local data directories for a TabletServer through `data.dirs`, distributing logs and KV data across several disks. When creating a new replica, the TabletServer chooses a disk with a lower load. This reduces the risk of a single disk becoming a capacity or I/O bottleneck and makes it easier to expand storage on a node.
+Fluss 1.0 supports configuring multiple local data directories for a TabletServer through `data.dirs`, distributing logs and KV data across several disks. When creating a new replica, the TabletServer chooses the least-loaded disk, balancing log and primary key replicas independently. This reduces the risk of a single disk becoming a capacity or I/O bottleneck and makes it easier to expand storage on a node.
```yaml
# Spread local data across multiple disks
@@ -405,8 +407,8 @@ As the first major release since Fluss graduated from the Apache Incubator, Flus
Get the release from the [Downloads page](https://fluss.apache.org/downloads/) and try the new capabilities with the [Quickstart](https://fluss.apache.org/docs/1.0/quickstart/flink/) and the [Gateway Real-Time Lakehouse Demo](https://fluss.apache.org/docs/1.0/quickstart/gateway-lakehouse/).
-Fluss 1.0 would not have been possible without the community. Thank you to the more than one hundred contributors below who participated in development, issue reporting, reviews, testing, documentation, and production validation:
+Fluss 1.0 would not have been possible without the community. Thank you to the contributors below who participated in development, issue reporting, reviews, testing, documentation, and production validation:
-> AlexZhao, Andrea Bozzo, Anton Borisov, Arnav-Panjla, Arnav\_Panjla, Aryamaan Singh, Beetle brank, CaoZhen, Chase Naples, Daeuk Choi, Dream95, Evan, Femi, ForwardXu, Gabriel Baldez, George Stamatakis, Gezi-lzq, Giannis Polyzos, Giovanny Gutiérrez, GuoYu, Hemanth Savasere, Hongshun Wang, Howie Wang, JB Onofré, Jackeyzhe, Jared Yu, Jared Yu (余启正), Jark Wu, Junbo Wang, Junfan Zhang, Kaiqi Dong, Kaixuan Duan, Karan Pradhan, Keith Lee, Kelvin Wu, Leonard Xu, Liebing, Liurnly, Lorenzo Affetti [...]
+> AlexZhao, Andrea Bozzo, Anton Borisov, Arnav-Panjla, Aryamaan Singh, Beetle brank, CaoZhen, Chase Naples, Daeuk Choi, Dream95, Evan, Femi, ForwardXu, Gabriel Baldez, George Stamatakis, Gezi-lzq, Giannis Polyzos, Giovanny Gutiérrez, GuoYu, Hemanth Savasere, Hongshun Wang, Howie Wang, JB Onofré, Jackeyzhe, Jared Yu (余启正), Jark Wu, Junbo Wang, Junfan Zhang, Kaiqi Dong, Kaixuan Duan, Karan Pradhan, Keith Lee, Kelvin Wu, Leonard Xu, Liebing, Liurnly, Lorenzo Affetti, Madhur Chandran, Mahesh [...]
Apache Fluss is under active development. Be sure to stay updated on the project, give it a try and if you like it, don’t forget to give it some ❤️ via ⭐ on [GitHub](https://github.com/apache/fluss).