The ScyllaDB team announces the release of ScyllaDB Open Source 2.3.2, a bugfix release of the ScyllaDB Open Source 2.3 stable branch. Release 2.3.2, like all past and future 2.3.y releases, is backward compatible and supports rolling upgrades.
Note that the latest stable release of ScyllaDB Open Source is ScyllaDB 3.0 and you are encouraged to upgrade to it.
- Get ScyllaDB 2.3.2 – Docker, binary packages (under “More Versions”)
- Get started with ScyllaDB
- Upgrade from 2.x.y to 2.x.z
- Upgrade from 2.3.x to 3.x
- Please let us know if you encounter any problems.
Issues solved in this release:
- iotune may report incorrect write bandwidth #4064
- After a Linux kernel upgrade, ScyllaDB may stop serving requests. This root cause is a bug in the Linux RWF_NOWAIT feature that can cause it to return EIO under some uncertain conditions, even though everything is fine. #3996
- In extreme out-of-memory situations, failure to allocate an internal data structure used for synchronizing memtable flushes can result in a segmentation fault and crash. #3931
- In some cases, ScyllaDB streaming failure stop the bootstrap of a new node joining the cluster #3732
- Wrong values if Size_estimates. Size_estimates is a field in a ScyllaDB system table which holds statistics on table size per node. Tools like Spark are using this field to calculate how large its reads should be #3916