Scylla Open Source 4.1.2

The ScyllaDB team announces Scylla Open Source 4.1.2, a bugfix release of the Scylla 4.1 stable branch. Scylla Open Source 4.1.2, like all past and future 4.x.y releases, is backward compatible and supports rolling upgrades.

These release notes also include defects fixed in Scylla Open Source 4.1.1.

Related links:

Issues fixed in Scylla 4.1.2:

  • CQL: Impossible WHERE condition returns a non-empty slice #5799
  • Docker: add option to start Alternator with HTTPS. When using Scylla Docker, you can now use --alternator-https-port in addition to the existing --alternator-port. #6583
  • UX: scylla-housekeeping, a service which check for the latest Scylla version, return “cannot serialize” error instead of propagating the actual error #6690
  • UX: scylla_setup: on RAID disk prompt, typing same disk twice cause traceback #6711
  • Stability: The concurrency control of the view update generation process was broken: instead of maintaining a constant fixed concurrency as intended, it could either reduce concurrency to almost 0, or increase it towards infinity #6774
  • Stability: Counter write read-before-write is issued with no timeout, which may lead to unbounded internal concurrency if the enclosing write operation timed out. #5069
  • Scylla crashes when CDC Log is queried with a wrong stream_id #6570

Issues fixed in Scylla 4.1.1:

  • CQL: Using LWT (INSERT ...IF NOT EXISTS) with two or more tables in a BATCH may cause boost::dynamic_bitset assertion #6332
  • Alternator: PutItem should output empty map, not empty JSON, when there are no attributes to return #6568
  • Alternator: missing UnprocessedKeys/Items attribute in BatchGetItem/WriteItems result #6569
  • UX: cql transport report a broken pipe error to the log when a client closes its side of a connection abruptly #5661
  • Storage: SSTable upgrade has a 2x disk space requirement #6682
  • Stability: Scylla freezes when casting a decimal with a negative scale to a float #6720
  • Alternator: modifying write-isolation tag is not propagated to a node of another DC #6513
  • Stability: Stream ended with unclosed partition and coredump #6478
  • Stability: oversized allocation when calling get token range API, for example during a Scylla Manager initiate repair #6297
  • Build: gdb cannot find shared libraries #6494
  • Stability: In a case when using partition or clustering keys which have a representation in memory which is larger than 12.8 KB (10% of LSA segment size), linearization of the large (fragmented) keys may cause undefined behavior #6637

23 Jul 2020