This short report brings to light some interesting commits to scylla-cluster-tests.git master from the last week.
Commits in the 89af4b3a…6fa21ca6 range are covered.
There were 15 non-merge commits from 8 authors in that period. Some notable commits:
Added a bit of documentation for loaders configuration parameters.
Introduced an enhanced script for copying an AMI to multiple regions, applying tags, and waiting for the image to become available. It detects if the image already exists before copying, supports dry-run mode, and can fetch the default region list from sct_config.py.
Continued the refactor of nemesis machinery: run_nemesis was moved from the Nemesis class to the node allocator abstraction. This change allows marking target nodes without requiring a Nemesis object and still preserves backward compatibility.
A performance test using latte for steady-state latency under a custom schema is now ready for regular triggering. It supports two flavors: one for vnodes and one for tablets.
Switched SCT test execution from unittest.main() to pytest, gaining better maintainability and access to pytest features like fixtures, subtests, and hooks.
Docker backend now uses the official ScyllaDB Docker image instead of rebuilding scylla-sct. A new DockerCmdRunner abstraction was added to run commands via Docker API instead of SSH.
scylla-bench was updated from version 0.2.4 to 0.2.5, fixing an issue where the -help flag was incorrectly printed to stderr instead of stdout.
See you in the next issue of last week in scylla-cluster-tests.git master!