The Scylla team is pleased to announce the release of Scylla 1.5.1, a bugfix release of the Scylla 1.5 stable branch. Note that the latest stable release is Scylla 1.6, you are encouraged to upgrade to it. Release 1.5.1, like all past and future 1.x.y releases, is backward compatible and supports rolling upgrades.
Related Links
- Scylla 1.5 release
-
Get started with Scylla 1.5 here.
-
Upgrade from 1.5.x to 1.6.y
-
Please let us know if you encounter any problems.
The following bugs were fixed in this release:
- Tracing: Starting a Node while other nodes send RPC messages requesting tracing causes an assert #1939
- Excessive disk usage during sstable resharding. Resharding will happen when importing SSTable from Cassandra or a Scylla server with a different core number #1952
- Error in Scylla install of scylla user already exist #1958
- Scylla setup script does not list local SSD disks, for example on a Google Cloud VM #1963
- scylla-housekeeping, a process checking for new Scylla release, failed if python-setuptools lib is not available #1884
- Prometheus: multiple metric families with the same name my hide Scylla metrics (not affecting scylla-grafana-monit
oring) #2077 - Prometheus: metric generation has quadratic behavior, affects latency #2078
- Scylla may crash under some circumstances when disk can’t keep up with writes for a long period: #2021
