mirror of
https://github.com/rustfs/rustfs.git
synced 2026-11-02 09:04:14 +00:00
hector
d2029ea8b0
The pf-testing environment is temporarily broken, so the nightly chain would end red in the performance lane. Take the chain through its designed skip path instead of deleting the lane: - prepare: comment out the pf-testing probe and force performance_ready=offline, so the lane job is skipped and the aggregate step passes --allow-skipped performance (marked TEMPORARY; restore the probe once the environment is healthy again). - summarize now forwards --allow-skipped; it previously recomputed the gate without it and would have stamped every nightly report complete=false during the outage. It also records skipped_lanes and renders them in the report. - the health collector validated complete evidence against the full 12-suite manifest; it now accepts exactly the designed skip (performance only) so nightly completes stay verifiable. - tests pin all three directions: skipped performance accepted, any other lane skipped or missing rejected, and prepare forcing offline. Verified locally: 63/63 unit tests, an 11-suite aggregate/summarize simulation, and the summarize CLI against a fixture evidence directory.