Mon, Nov 2, 2020

TiDB Monthly Update for October 2020

Releases

We have three releases this month:

TiDB 4.0.8 Release Notes
- New features
  - Support the new aggregate function APPROX_PERCENTILE
  - Support pushing down CAST functions
  - Support the snapshot-level consistent replication
- Improvements
  - Prioritize low-selectivity indexes in the greedy search procedure of Selectivity() in TiDB
  - Add the Fast-Tune panel page to assist performance diagnostics in TiKV
  - Add monitoring metrics of Raft logs in TiFlash
- Bug Fixes
TiDB Data Migration 2.0 GA Release Notes
- Improvements
  - Optimize the safe-mode setting to ensure the eventual consistency of data when the upstream database, such as Amazon Aurora and Aliyun RDS, does not support FTWRL in the full export #981 #1017
  - Support automatic configuration of sql_mode for data migration based on the global sql_mode of upstream and downstream databases and sql_mode of binlog events #1005 #1071 #1137
  - Support automatic configuration of max_allowed_packet from DM to the downstream TiDB, based on the global max_allowed_packet value of the downstream TiDB #1071
- Bug fixes
TiDB Operator 1.1.6 Release Notes

Blog posts

Yihao Zhang, the Senior development engineer at Xiaohongshu, wrote an article How We Use a Scale-Out HTAP Database for Real-Time Analytics and Complex Queries that described why they chose TiDB and how TiDB's real-time HTAP capabilities help them manage their data in some scenarios.

In Lessons from TiDB's No. 1 Bug Hunters Who've Found 400+ Bugs in Popular DBMSs, Manuel Rigger, the postdoctoral fellow at ETH Zurich, described the techniques that have made him and his colleague, Professor Zhendong Su, TiDB's #1 bug hunters.

How We Use a Distributed Database to Achieve Horizontal Scaling Without Downtime, written by Zhendong Chen, the software development engineer at Bank of Beijing, shared with you why they chose TiDB and how they use it to achieve horizontally scalable, always-on infrastructure.

Yu Han, the engineer at Information Technology Operations Center, Bank of China, wrote an article How Bank of China Uses a Scale-Out Database to Support Zabbix Monitoring at Scale that introduced the traditional Zabbix monitoring solution they adopted and their pain points at the time, how they used TiDB in Zabbix, and their further plans to optimize TiZabbix.

We published TiDB X Hacktoberfest 2020 - An Invitation to Open Source to invite you to be part of us, starting from our handpicked issues with proper mentoring and assistance along your journey.

We also published TiCDC GA: Offering High-Availability Replication Services for Production Environments to announce the general availability of TiCDC, and to walk you through TiCDC's features, application scenarios, and real-world case studies.

Youzhi Zhu, the big data architect at ZTO Express, wrote an article Why We Migrated from Exadata to a Scale-Out HTAP Database for Near Real-Time Analytics, described why they chose TiDB as their solution and how TiDB helps scale out their database and supports their multi-dimensional analytics with query response times in minutes.

In Making an HTAP Database a Reality: What I Learned from PingCAP's VLDB Paper, Xianlin Chen, the DBA at PalFish, shared with you his thoughts about TiDB's implementation of an HTAP database that provides strong data consistency and resource isolation, as well as his expectations of TiDB in the future.

Running a Scale-Out Database on ARM as a MySQL Alternative, written by Birong Huang, the senior engineer at U-Next, shared with you why they chose to put their data in TiDB and run it on ARM, and the detailed benchmarking tests of TiDB on ARM.

Our engineer Shuang Chen wrote an article Metrics Relation Graph Helps DBAs Quickly Locate Performance Problems in TiDB to introduce a new feature in its web UI TiDB Dashboard: the metrics relation graph.