QuestDB
Day 32 after signup
QuestDB 10.2 improves power-loss durability
Power-loss durability for much less than sync, resent rows that no longer overwrite the first copy, and Broadridge's CPO on why they chose QuestDB.
Why it works
- Lead with the release’s most important change and explain its practical benefit before listing secondary improvements.
- Pair the release announcement with upgrade guidance and one clear next step, then place related customer and technical content afterward.
How it is built
- Release headline
- Lead feature explanation
- Additional improvements
- Upgrade guidance and download
Email copy
QuestDB 10.2: power-loss durability without the cost of sync QuestDB 10.2 is out. The headline is a new adaptive commit mode. It makes commits on WAL tables survive an operating system crash or a power loss, for much less than sync costs. Each commit flushes only its write-ahead log files, and the table files are flushed once a minute. The default is unchanged, and adaptive is opt-in with one setting (docs). 10.2 also brings: A filtered copy of a table, kept up to date as rows arrive, with passthrough views. Views that only keep the rows still relevant, such as outages in effect or the latest reading per device, with EXPIRE ROWS. Resent rows that no longer overwrite the copy that arrived first, with DEDUP SKIP. Detail rows, subtotals and a grand total from one query and one scan, with GROUPING SETS, ROLLUP and CUBE. Check the breaking changes before upgrading. The release post has all the details. Download QuestDB Hear from Broadridge's CPO why they chose QuestDB At STAC Summit New York in October, our CEO Nicolas Hourcard sat down with Brian Pomraning, Chief Product Officer at Broadridge, to talk about data infrastructure for trading without an end of day, and why Broadridge chose QuestDB. The recording is now online. Watch the fireside chat Brian Pomraning of Broadridge and our CEO Nicolas Hourcard at STAC Summit NYC. More reading: Replaying a full trading day: 354 million US equity top-of-book events streamed back from QuestDB as Arrow, in strict time order in about two minutes, or in 32.5 seconds over parallel connections. Code and hardware are included so you can rerun it. Market data replay for backtesting No more nightly Parquet export: QuestDB SQL, Polars and DuckDB read the same Parquet partitions of one table and return the same numbers. One table, three readers Port congestion from free AIS data: a community post on how Port Context Engine tracks ships waiting at four European ports with NATS and a QuestDB container capped at 2 CPUs and 1 GB. Measuring port congestion with NATS and QuestDB Questions or feedback? Join our public channels. Slack | Community Forum | LinkedIn | X Copyright © 2026 QuestDB Technology Inc. London Office: 8 Devonshire Square, London EC2M 4YJ, UK New York Office: 575 Lexington Ave, New York, NY 10022, USA You received this because you subscribed to QuestDB's newsletter. Unsubscribe