[RELEASE] ScyllaDB 5.2 RC5

Release Notes
, , ,
1

The Scylla team is pleased to announce ScyllaDB Open Source 5.2 RC5, a Release Candidate for Scylla Open Source 5.2 minor release.

We encourage you to run ScyllaDB 5.2 release candidates on your test environments; this will help ensure that an upgrade to ScyllaDB 5.2 General Availability will proceed smoothly with your workload. Use the release candidate with caution; RC5 is not production-ready yet. You can help stabilize Scylla Open Source 5.2 by reporting bugs here.

Only the last two minor releases of the ScyllaDB Open Source project are supported. Once Scylla Open Source 5.2 is officially released, ScyllaDB Open Source 5.2 and 5.1 will be supported, and ScyllaDB 5.0 will be retired.

For a complete description of ScyllaDB 5.2 see https://forum.scylladb.com/t/release-scylla-5-2-rc1/330.

Related Links

Updates and bug fixes since 5.2 RC4 (not including tests, docs updates)

  • Stability: Adding nodes to a large cluster (90+ nodes) may cause existing nodes to crash. The root cause is quadratic behavior in get_address_ranges function #12724

  • Stability: a rare issue when updating the schema more than once in once per millisecond #13594

  • Stability: Compaction manager “periodic reevaluation” is one-off. The Compaction manager intended to periodically reevaluate compaction needs for each registered table. But it’s not working as intended. The reevaluation is one-off. This means that compaction was not kicking in later for a table, with low to none write activity, that had expired data 1 hour from now. #13430

  • Stability: Internal error in a COUNT request with empty IN. The query “select count(*) from {table1} where p in ()” should result in the count 0, because the empty p in () matches no row. However, what we get in Scylla now is an internal error. #12475

  • Stability: a very rare failure in a mutation validation failure test, when Materialized View building stops just as it reaches the end of the base table, deciding it needs to stop and also reset the reader to the start token at the same time while the compaction state also thinking the last read partition will be resumed. #12629

  • Tools: total disk space used metric incorrectly tells the amount of disk space ever used, which is wrong. It should tell the size of all SSTable being used plus the ones waiting to be deleted. Live disk space used shouldn’t account for the ones waiting to be deleted, and live SSTable Count shouldn’t account SSTable waiting to be deleted. #12717

  • Stability: reader_concurrency_semaphore: inactive reader eviction is too aggressive, leading to Reads timing out, especially in range scans. #11803

  • Stability: scylla hangs on shutdown in test_total_space_limit_of_commitlog dtest #12810

  • Stability: Segmentation fault upon wrong named bind markers passed by driver (identify in Rust drivers tests) #12727

  • Performance: SSTable set is left uncompacted post off-strategy compaction completion, which may lead to suboptimal read and space amplification until the next compaction #13429

  • CQL: TTL unexpected behavior when setting to 0 on a table with default_time_to_live #6447

  • CQL: UDA keep using old UDF even after the UDF is replaced #12709 UDF and UDA are experimental in Scylla 5.2

  • Stability: a rare crash due to null pointer dereference: clear_gently of disengaged unique_ptr dereferences nullptr #13636