What rqlite/rqlite shipped
Written by FoxPlug from the repository's own releases; not affiliated with Rqlite. Paste your own repository in the box below and FoxPlug writes the same weekly update for your product. This page is an automatic summary of the releases, pull requests and commits in github.com/rqlite/rqlite; Rqlite did not write it and does not use or endorse FoxPlug. Every line links to the change it describes.
Get a weekly update like this for your product, free
Week of October 5, 2026
What shipped
- A non-voting node can now be promoted to a voter while keeping the address it already has in the cluster configuration. Pull request #2868
- A voting node can now be demoted to a non-voter, staying in the cluster as a read-only replica. Pull request #2867
- The CDC service now always answers snapshot sync requests, so the Store no longer blocks snapshotting until its sync timeout expires when the flush marker write fails. Pull request #2870
- The CDC high watermark now advances past dropped batches, so a batch that exhausts its retries is no longer resent after a restart or loss of leadership. Pull request #2873
- A follower's CDC high watermark no longer moves backwards, so a node that delivered events as leader and then stepped down will not accept a lagging watermark from the new leader. Pull request #2872
- The Raft Snapshot threshold is now checked every 5 seconds by default instead of every 10 seconds. Pull request #2845
- Bound TEXT and BLOB parameter lengths now count toward the compression size threshold in RequestMarshaler.Marshal. Pull request #2849
- The API now supports choosing which error style to return. Pull request #2864
- Gzip memory usage is now optimized using a sync.Pool. Pull request #2869
- CDC stats now returns promptly on a closed queue. Pull request #2871
Why it matters
This week's changes make cluster membership more flexible: nodes can be promoted to voters and demoted to non-voters without changing their addresses. Several CDC fixes close gaps where dropped or lagging batches could be resent or block snapshotting. The default Raft Snapshot threshold is also checked more often, which matters under high write load.
Changelog entry
- Support promoting a non-voting node to a voter. Pull request #2868
- Support demoting a node from voter to non-voter. Pull request #2867
- Cdc: always answer snapshot sync requests. Pull request #2870
- Advance CDC high watermark past dropped batches. Pull request #2873
- Never move a follower's CDC high watermark backwards. Pull request #2872
- CDC stats returns promptly on closed queue. Pull request #2871
- Check Raft Snapshot threshold every 5 seconds by default. Pull request #2845
- Count bound TEXT and BLOB bytes toward the compression size threshold. Pull request #2849
- Support choosing which error style to return. Pull request #2864
- Optimize gzip memory usage via sync.Pool. Pull request #2869
- Consolidate console home and topology. Pull request #2876
- Fix off-by-one error in CDC retry logic. Pull request #2856
This week: promote non-voting nodes to voters, demote voters to non-voters, several CDC fixes for dropped and lagging batches, snapshot threshold checked every 5 seconds, and gzip memory optimized with sync.Pool.
This week in rqlite: non-voting nodes can be promoted to voters, and voting nodes can be demoted to non-voters while staying in the cluster as read-only replicas. Several CDC fixes landed, including advancing the high watermark past dropped batches, keeping a follower's high watermark from moving backwards, and always answering snapshot sync requests so snapshotting is not blocked. The Raft Snapshot threshold is now checked every 5 seconds by default, and gzip memory usage is optimized with sync.Pool. Bound TEXT and BLOB bytes now count toward the compression size threshold.
Week of September 28, 2026
What shipped
- V10.5.2 released with autocheckpointing disabled at connect time to prevent driver timeouts from creating uncontrolled connections. Release
- CDC events now emit for every request in bulk operations, improving change data capture granularity. Pull request #2840
- CDC streamer index no longer clears after Commit or Rollback, preserving state across transactions. Pull request #2839
- New Cluster tab added to the web UI for better cluster visibility. Pull request #2835
- Node-local SQLite errors now trigger exit instead of continuing, preventing cluster divergence from transient failures. Pull request #2831
- Queued-writes retries are now configurable for better control over write behavior. Pull request #2827
- CDC high water mark updates can now require credentials for enhanced security. Pull request #2826
- Stale read check improved to handle blackholed AppendEntries RPC messages. Pull request #2834
- Driver timeout handling improved by disabling autocheckpointing on connection creation. Pull request #2841
- DB Dump() method now returns the number of bytes written. Pull request #2828
Why it matters
This week focused on stability improvements around change data capture, node resilience, and error handling. Critical fixes for SQLite error handling prevent nodes from diverging in the cluster, while CDC enhancements ensure reliable capture of all write operations. Configuration additions give operators more control over retry behavior.
Changelog entry
- V10.5.2: Autocheckpointing disabled at connect time Release
- Disable autocheckpointing when driver creates connection after timeout Pull request #2841
- Emit CDC events for every request in bulk operations Pull request #2840
- CDC streamer index preserved across Commit and Rollback Pull request #2839
- Exit on node-local SQLite errors to prevent cluster divergence Pull request #2831
- Queued-writes retries now configurable Pull request #2827
- CDC high water mark updates can require credentials Pull request #2826
- New Cluster tab in web UI Pull request #2835
- Stale read check handles blackholed AppendEntries RPC Pull request #2834
- AppendEntries reception time recorded on each node Pull request #2832
- DB Dump() returns number of bytes written Pull request #2828
- UI terminology updated to use 'read replicas' Pull request #2836
V10.5.2 is out. Better CDC event handling, improved autocheckpointing at connect time, and more resilient cluster behavior when nodes encounter SQLite errors.
V10.5.2 released with improvements to change data capture reliability and cluster stability. Node-local SQLite errors now trigger exit to prevent cluster divergence, CDC events emit for each bulk request, and autocheckpointing is disabled at connection creation. New Cluster UI tab and configurable queued-writes retries give operators better visibility and control.
Week of September 21, 2026
What shipped
- V10.3.6 released. Release
- Database Backup now returns the size of the backup file. Pull request #2824
- Backup calls return the number of bytes written. Pull request #2822
- Snapshot ordering is kept correct when the system clock moves backwards. Pull request #2812
- Rename and Remove operations are now retried. Pull request #2817
- Generation field removed from snapshot names. Pull request #2819
- Rename and Remove operations added with retry logic. Pull request #2814
- Standard library replaced with fsutil. Pull request #2816
- Testing against minio removed due to pricing changes. Pull request #2823
Why it matters
Backup operations now provide detailed information about written bytes and file size, making it easier to verify backup success. The fix for clock skew ensures snapshots maintain correct ordering even in edge cases. Retry logic for file operations improves reliability in environments with transient failures.
Changelog entry
- Database Backup returns size of backup file Pull request #2824
- Backup calls return number of bytes written Pull request #2822
- Keep snapshot ordering correct when the clock moves backwards Pull request #2812
- Rename and Remove are retried Pull request #2817
- Remove generation from snapshot names Pull request #2819
- Replace standard library with fsutil Pull request #2816
Rqlite v10.3.6 is out. Backup operations now report bytes written and file size. Snapshot ordering is fixed when the system clock moves backward. Rename and Remove operations are retried for better reliability.
Rqlite v10.3.6 is available. This release improves backup operations by reporting bytes written and backup file size, fixes snapshot ordering when system clocks move backward, and adds retry logic to Rename and Remove operations for better fault tolerance in distributed environments.
Week of September 14, 2026
What shipped
- V10.3.5 released. Release
- Configurable slow-query logging added with -query-log flag supporting stderr or JSON file output and minimum duration thresholds. Pull request #2768
- Snapshot IDs now include generation information. Pull request #2807
- Snapshots can be filtered by term and index. Pull request #2810
- PRAGMA syntax validation strengthened to reject alternate forms that could bypass disallowed-PRAGMA checks. Pull request #2794
- CDC event handling fixed to discard rolled-back events from failed autocommit inserts. Pull request #2791
- HTTP logic error for queued statements corrected. Pull request #2802
- HTTP last backup time access synchronized for thread safety. Pull request #2799
- Dependencies upgraded. Pull request #2811
Why it matters
This week's releases include improvements to query logging observability, snapshot filtering capabilities, and several critical bug fixes for CDC events, PRAGMA validation, and HTTP operations. These changes enhance debugging capabilities and system reliability for distributed deployments.
Changelog entry
- Added configurable slow-query logging via -query-log flag with support for stderr or JSON file output and minimum query duration thresholds Pull request #2768
- Snapshot IDs now contain generation information Pull request #2807
- Snapshots can be filtered by term and index Pull request #2810
- Strengthened PRAGMA validation to reject alternate syntax forms including parenthetical notation that could bypass checks Pull request #2794
- Fixed CDC event handling to discard rolled-back preupdate events from failed autocommit inserts Pull request #2791
- Fixed HTTP logic error for queued statements Pull request #2802
- Synchronized HTTP last backup time access for thread safety Pull request #2799
- Set Served by header before first backup byte Pull request #2800
- Set DB load response time for all formats Pull request #2801
- Added handling for negative trailing logs Pull request #2798
- Fixed UI console issues Pull request #2804
- Tweaked query logging Pull request #2805
- Changed DriverConfig to use pointer semantics Pull request #2806
- Upgraded dependencies Pull request #2811
Rqlite v10.3.5 is out. Configurable slow-query logging, snapshot filtering, stronger PRAGMA validation, and CDC event fixes improve observability and reliability.
Rqlite v10.3.5 released with configurable slow-query logging for SQL observability, snapshot filtering by term and index, reinforced PRAGMA syntax validation, and fixes for CDC event handling in failed transactions. These improvements enhance debugging, filtering, security, and distributed database reliability.