Update changelog.md

about/changelog.md

@@ -9,6 +9,46 @@ products: [cloud]
 
 All the latest features and updates to $CLOUD_LONG.
 
+## New navigation in Console
+<Label type="date">December 5, 2025</Label>
+
+We have updated the design of the navigation in Console to improve consistency, reduce distractions, and make it easier and faster to navigate through the different menus.
+
+![New console navigation](https://assets.timescale.com/docs/images/console-new-navigation.png)
+
+## S3 connector GA and Tiger Lake public beta
+<Label type="date">November 28, 2025</Label>
+
+### S3 source connector general availability
+
+The S3 source connector is now production-ready, delivering major improvements in reliability, performance, and correctness across the entire ingestion pipeline. We resolved extensive issues related to file state transitions, workflow ordering, live sync consistency, and error handling, while adding support for retries, skipping files, better conflict handling, and more stable pause/resume behavior. Import operations are now more resilient, deterministic, and traceable, with clearer progress reporting and more accurate file and worker state visibility.
+
+Performance and scalability have been significantly enhanced through better autoscaling, improved worker resource allocation, faster preview and import operations, optimized scheduling, reduced memory usage, and higher throughput for small and large files. 
+
+The user experience has been upgraded across the UI with improved file filtering, pagination, column mapping, input validation, clearer states and sizes, and more robust multi-step flows. Numerous bugs affecting import previews, table selection, schema handling, hypertable creation, and navigation were fixed. Combined, these changes make the S3 source connector significantly more stable, predictable, and user-friendly—ready for GA adoption.
+
+### Tiger Lake public beta
+
+Tiger Lake is now available in public beta and ready for broader use. The public beta includes:
+
+**Full DML support**
+* Replication of INSERT, UPDATE, and DELETE operations on hypertables and regular Postgres tables to Iceberg.
+
+**High-performance ingestion**
+* Decoupled CDC and full table import pipelines.
+* CDC ingest throughput: ~30,000 records/second.
+* Initial full table import throughput: ~300,000 records/second.
+
+**Enhanced resilience and self-healing**
+* Automatic recovery when a replication slot disappears—all data is correctly replayed, keeping Iceberg tables eventually consistent.
+* Seamless recovery in case of hardware failure without restarting ingestion from scratch.
+
+**Improved deployment experience**
+* Validation of ARNs before service deployment to prevent misconfiguration.
+* Compatibility against hypertables with data in rowstore and columnstore.
+
+Read the [documentation](https://www.tigerdata.com/docs/use-timescale/latest/tigerlake) to get started.
+
 ## TimescaleDB v2.23 – improved getting started, automatic columnstore, and faster continuous aggregate invalidation tracking
 <Label type="date">November 14, 2025</Label>