💫 New Features & Changes
- Added an autoscaling annotation to choose a different aggregation algorithm for the autoscaling metrics. This is experimental currently. (#10840, @vagababov)
- Added
autocreateClusterDomainClaimsflag to network config map. (networking#330, @julz)
🐞 Bug Fixes
- Adds validation that a default max-scale is set if a max-scale-limit is specified in the autoscaler configmap (since otherwise the default max-scale, i.e. 0 = no max, would fail validation as it is above the max-scale-limit). (#10921, @julz)
- Bumped the resource request and limits of the autoscaler to 100m/100Mi, 1000m/1000Mi respectively. (#10865, @markusthoemmes)
- Fixed a regression where the pod bringup time might have a latency of 10s or more even though the container should be up quickly. (#10992, @markusthoemmes)
- Reduced the necessary memory allocations in the activator significantly, especially with disabled tracing. (#11016, #11013, #11009, #11008, @markusthoemmes)
- Fix the incorrect Gateway name format for DomainMapping auto TLS feature for net-istio implmenetation. (net-istio#532, @ZhiminXiang)
Dependencies
Added
Nothing has changed.
Changed
- github.com/google/go-cmp: v0.5.4 → v0.5.5
- github.com/prometheus/common: v0.15.0 → v0.19.0
- go.opencensus.io: v0.22.6 → v0.23.0
- google.golang.org/grpc: v1.35.0 → v1.36.0
- knative.dev/caching: 1212288 → 5691bb3
- knative.dev/hack: 8368e1f → b6ab329
- knative.dev/networking: 088986a → 999a770
- knative.dev/pkg: 84c98f3 → 952fdd9
Removed
Nothing has changed.