Releases: kedify/charts
Releases · kedify/charts
kedify-agent: v0.5.1
Chart v0.5.1 CHANGELOG:
- Update CRDs for Kedify (#415, #413, #412)
- Update Helm Docs (#407)
- agent image version bump to v0.5.1
Image v0.5.1 CHANGELOG:
Features:
- mc: parallel member cluster health check
- Narrow the set of metrics that agent collect from /metric endpoints to only those that we actually use
- dsj: Priority based Failover
- dsj: set 'keepAll' duplication handling policy as default
- dsj: Rename rebalancingPolicy -> failoverPolicy
Fixes:
- dsj: improvements for status reporting and handling of cluster filter
- Change grpc reconnect logic on network failure
kedify-agent: v0.5.0
Chart v0.5.0 CHANGELOG:
- Update CRDs for Kedify (#409)
- kedify-agent: bump deps; KEDA v2.19.0-0 (#408)
- chore(deps): update kedify-predictor docker tag to v0.1.4 (#402)
- agent image version bump to v0.5.0
Image v0.5.0 CHANGELOG:
Features:
- dsj: metrics for job counters and reduce event spam for happy path
- HA mode
- switchable reconcilers - multicluster DSO and DSJ by default disabled - enable through feature flags
Fixes:
- dsj: Fix status reporting of desired job count
Chore:
- mc: refactor multicluster dsj and dso shared code
keda: v2.19.0-0
Chart v2.19.0-0 CHANGELOG:
- keda image version bump to v2.19.0-0
Upstream releases
kedify-agent: v0.4.15
Chart v0.4.15 CHANGELOG:
- Update CRDs for Kedify (#399)
- on/off for various agent controllers (#398)
- agent image version bump to v0.4.15
Image v0.4.15 CHANGELOG:
Features:
- DSJ: add per-cluster JobSpec overrides (nodeSelector, tolerations, affinity, image, env, resources)
- metrics: expose per-cluster dso/dsj health and KedifyConfiguration multicluster state
- Implement pause mechanism via annotation on DSO/DSJ
- Run multicluster e2e tests on PRs
- dsj: Tweak default failover parameters
Fixes:
- Regenerate manifests for PRP kubebuilder tags
- Allow image workflow publish packages
kedify-agent: v0.4.14
Chart v0.4.14 CHANGELOG:
- Update CRDs for Kedify (#391, #388, #387)
- add PodResourceAutoscaler (#390)
- agent image version bump to v0.4.14
Image v0.4.14 CHANGELOG:
Features:
- mc: status visibility and condition improvements
- mc: events for dsj and dso to improve troubleshooting
- dsj: refine ready condition when unable to establish gRPC connection
- dsj: rename quarantine to taint and apply soft taint
- feat: Add PodResourceAutoscaler (PRA)
Fixes:
- mc: fix DistributedScalingJob scaling strategies
- mc: fix cluster ID/name mismatch in cluster filtration for DistributedScaledJob
- dsj: allow grace period before re-scheduling pending pods
kedify-agent: v0.4.13
Chart v0.4.13 CHANGELOG:
- Update CRDs for Kedify (#385)
- agent image version bump to v0.4.13
Image v0.4.13 CHANGELOG:
Features:
- Scrape metrics also for predictor & OTel scaler
- Improve insight into PRP controller with Metrics & Events
- multi-cluster: member periodic health checks recorded in KedifyConfiguration status
kedify-agent: v0.4.12
Chart v0.4.12 CHANGELOG:
- Add observedHelmValues & helmReleaseName fields to kedifyconfig status (#379)
- chore(deps): update kedify-predictor docker tag to v0.1.1 (#378)
- kedify-agent: bump deps; KEDA v2.18.3-6 (#377)
- chore(deps): update kedify-predictor docker tag to v0.1.0 (#375)
- agent image version bump to v0.4.12
Image v0.4.12 CHANGELOG:
Features:
- Update CRDs
- Add observed Helm values to kedify status
- scrape also limits&requests for initContainers and ephemeralContainers
keda: v2.18.3-6
Chart v2.18.3-6 CHANGELOG:
- keda image version bump to v2.18.3-6
Image v2.18.3-6 CHANGELOG:
Changes:
- fix predictive scaler (reporting wrong value as metric)
kedify-agent: v0.4.11
Chart v0.4.11 CHANGELOG:
- kedify-agent: bump deps; KEDA v2.18.3-5 (#373)
- kedify-agent: bump deps; kedify-proxy v0.0.10 (#371)
- chore(deps): update kedify-predictor docker tag to v0.0.6 (#364)
- agent image version bump to v0.4.11
Image v0.4.11 CHANGELOG:
Features:
- automatically determine namespace for kedify-proxy fleet