[RELEASE] ScyllaDB 2025.1.17

The ScyllaDB team is pleased to announce the release of ScyllaDB 2025.1.17, a production-ready patch release for ScyllaDB 2025.1 Long Term Support (LTS) Release.

We recommend upgrading to the next Long Term Support (LTS) release 2026.1 to get the latest features.

More information on ScyllaDB Version Support Policy is available here.

Related Links

Bug Fixes

The following issues are fixed in this release.

Audit

  • Audit writes used timeouts of up to one hour. If a replica did not respond to an audit write (for example, during a rolling upgrade), topology operations could stall for that long, and a joining node could remain in the bootstrap state. Audit writes now use the same five-second timeouts as tracing writes, so topology operations continue promptly.
    scylladb#32583

Compaction

  • When a row had a shadowable tombstone and a regular tombstone with timestamps in different time windows, the timestamp-based splitting writer wrote both into the same time window. This could cause Time Window Compaction Strategy (TWCS) reshape to repeat endlessly. Each tombstone part is now written to the time window that matches its own timestamp, so reshape finishes as expected.
    scylladb#32214

  • Shutting a node down very early during startup is handled more gracefully. The compaction task-manager module is now always unregistered when the compaction manager stops, even if compaction was not yet enabled, so a fast startup-then-shutdown sequence completes cleanly.
    scylladb#32390

  • If a node shut down early during startup, before compaction was enabled, a late disk-space-monitor callback could re-enable the compaction manager while it was stopping. Shutdown now moves the compaction manager into its stopped state, so a node stops cleanly after an early startup error.
    scylladb#32390

  • When a keyspace was dropped shortly before a node was stopped, an error during shutdown could interrupt the schema change and, in some cases, cause the node to restart. Nodes now shut down cleanly in this scenario.
    scylladb#32390

CQL

  • A SELECT that combined a tuple range restriction on clustering columns with filtering could return wrong results when those clustering columns were declared with DESC (descending) order. Without a secondary index the query returned no rows; with a secondary index it could return exactly the rows that should have been excluded. Such queries now return the correct rows regardless of the clustering order.
    scylladb#31797

  • When a secondary index was temporarily out of sync with its base table, for example because of replication lag, a query that used the index could return rows that did not match the WHERE clause. ScyllaDB now checks every row read through an index against the base row before returning it, so indexed queries always return results that match their predicates. With this change, indexed queries with a single predicate use the same filtering path as indexed queries that already had more than one predicate.
    scylladb#31994

  • SELECT ... FROM MUTATION_FRAGMENTS(<table>) ignored GROUP BY and returned one row aggregated over all fragments instead of one row per group. Aggregates over more than one internal page also came back as several partial rows. For example, a COUNT over 25,000 fragments returned 10000, 10000, and 5000. Aggregate and GROUP BY queries on MUTATION_FRAGMENTS() now page and group correctly and return the expected results.
    scylladb#32039

  • A query that used a bind variable inside an aggregate’s argument (for example, max(intasblob(?))) or applied to its result (for example, max(m)[?]) could cause the node to restart. Aggregates are now evaluated with the query options, so such queries return the expected results, with or without GROUP BY.
    scylladb#32952

Maintenance

  • Maintenance tasks could return a failure status even if a table was successfully dropped during the operation. The tasks service now correctly returns a successful status if the underlying table was dropped. This provides accurate feedback to maintenance tools about the task outcome.
    scylladb#29645

Monitoring

  • Tablet repair logged several per-shard and per-request messages at the info level, which could flood the system journal and cause other log messages to be suppressed. These messages are now logged at the debug level, while messages that report an actual hints or batchlog flush stay at the info level, reducing log volume during tablet repair without losing relevant information.
    scylladb#32927

Networking

  • With RPC compression enabled, an internode message could, in rare cases, be decompressed incorrectly and rejected with an RPC checksum error, because the LZ4 decompression ring buffer was slightly too small. The buffer now includes the margin required by LZ4, so compressed internode messages are always decoded correctly.
    scylladb#32670

Repair

  • Repair-based node operations now use topology guards, as streaming-based operations already do. Late messages from an already-completed operation are now rejected instead of being mistaken for part of a newly started one, which improves the safety of node operations that involve writes.
    scylladb#27759, scylladb#28296

  • A non-numeric value for the ranges_parallelism repair option was silently treated as 0, which caused the repair to hang forever. Invalid and out-of-range values (below 1) are now rejected immediately with a clear “bad parameter” error instead of hanging.
    scylladb#31810

  • Before a repair, a node replays pending hints and batchlog entries. Batchlog replay had no time limit, so if it stalled, every later repair on that node was blocked until the node was restarted. Batchlog replay now has a timeout, like hints replay, so later repairs can proceed.
    scylladb#32471

Service Levels

  • The computation of effective service levels now gracefully handles cycles in the role grants graph. Under normal operation the role grants graph is acyclic, but inconsistent state migrated from pre-Raft authorization or manual edits could introduce cycles that caused the previous topological sort algorithm to fail with an exception, breaking service level updates on the affected node. The algorithm is replaced with a depth-first search with memoization that detects and reports cycles via a warning log while still computing a usable set of effective service level parameters.
    scylladb#30629

Stability

  • Some internal queries in system tables, such as lookups in system.repair_tasks and system.topology_requests, put parameter values directly into the query text. Each distinct value was cached as a separate prepared statement that was never evicted, so memory use grew steadily over time on long-running nodes. These queries now use bound parameters, which keeps the internal statement cache bounded.
    scylladb#32124

  • An error while processing a gossip update (for example, under memory pressure) could leave a node with incomplete information about another node, which gossip did not correct afterwards. Incomplete updates are now discarded and the full state is received again, so gossip state stays consistent across the cluster.
    scylladb#32431

  • On hosts with an uneven NUMA node layout, a node started with a reduced number of shards (--smp) could fail at startup. Such nodes now start normally.
    scylladb#32464

Tablets

  • If a tablet rebuild operation was rolled back, any missing replicas were not properly added back into the topology. The service now adds missing replicas if a tablet rebuild was rolled back. This maintains the correct replication factor and prevents data unavailability following a rollback scenario.
    scylladb#31948