Hi all, Following suggestions from the ASF Board to broaden the scope of our regular updates and present them as community-oriented rather than company-specific, we are resuming this series under the title “Weekly Update of Timecho’s Work to the Community.”
These updates summarize Timecho’s recent open-source contributions to Apache IoTDB and related projects, as well as community events and activities - if there were some. [IoTDB Development Tracker]: - Completed development of the approximate quantile function. - Enabled streamed result return and memory control for fetchSchema. - Optimized last(s1) and last_by(time, s1) in table mode to leverage latest-value caching for acceleration. - Enhanced error messages for table mode INSERT when the number of values does not match the number of columns. - Fixed a potential path traversal vulnerability in the IoTConsensus snapshot loading interface. - Resolved memory inflation issues in IoTConsensus batch processing on the receiver side. - Improved error handling for password history creation failures and for changing passwords of non-existent users. - Added default compression configuration options for different data types. - Fixed getTimeStamp in the C++ client to support non-first columns. - Addressed uncaught exceptions when creating a TsFile directory fails. - Fixed NPEs caused by multiple errors during processing. - Resolved several Pipe-related issues: (1) forkJoinPool concurrency problems. (2) Client thread leaks after DROP PIPE. (3) Incorrect Remaining count statistics. (4) Memory leaks due to frequent logins by Pipe receivers. (5) TsFile resource leaks. (6) Unfair locking in Sink Subtasks blocking start threads. (7) Adjusted CN CREATE PIPE rejection strategy for low-memory situations. - Fixed issue where ADD COLUMN to a non-existent view caused errors. [Calls for Community Participation]: Table mode adaptation for Telegraf, DataEase, Grafana, DataX, and ThingsBoard. [Community & Events] On July 5, 2025, we co-hosted the *Time Series Database Tech Innovation Summit 2025* with Tsinghua University. - Theme: “Next Frontier: DB + AI”. - The event brought together developers, researchers, and end users to exchange insights on time-series infrastructure, real-world applications, and future trends. - Several talks shared recent developments, practical use cases, and best practices with Apache IoTDB, covering topics such as industrial deployment, AI integration, and ecosystem tools. - Video recordings of selected sessions will be published gradually on our Youtube Channel. [1] [1] https://youtu.be/rMTbLGGcsbM?si=gjcz53Ycq_UC9-Ef Wish all a nice weekend! Best regards, Pengcheng The Timecho Team
