Polycrate API 0.11.22 released: LoadBalancer Metrics Fix
Polycrate API 0.11.22 fixes the LoadBalancer metrics query. The detail UI now shows correct …
With Polycrate API 0.11.16, we address two critical bugs that could occur in production environments.
The Issue: S3 Buckets remained permanently in the DEGRADED status and were never created at the provider (MinIO). The logs showed no error—only a misleading info message about missing metrics.
The Cause: During the migration to run_reconciliation() in 0.11.12, early returns were introduced, skipping the REPAIRS section. For new buckets without metrics, create_bucket() was never called.
The Solution: Metrics fetching no longer blocks the repair process. All early returns have been removed and replaced with logging.
Post-Update: Trigger affected S3 Buckets via the Reconcile button.
The Issue: LoadBalancers showed 11.4 TB of traffic on “30 Days” even though they were only 6 hours old. Several LBs displayed identical values.
The Cause: Fallback queries without label filters summed all HAProxy metrics in the cluster. Additionally, Prometheus extrapolated the values unrealistically high.
The Solution: Removed fallbacks, used instant queries, and limited time periods to the LBI lifespan.
The polycrate-api block has been updated to version 0.5.20:
polycrate pull cargo.ayedo.cloud/ayedo/k8s/polycrate-api
polycrate run polycrate-api install
Polycrate is ayedo’s Infrastructure-as-Code platform for declarative multi-cluster management. Learn more →
Polycrate API 0.11.22 fixes the LoadBalancer metrics query. The detail UI now shows correct …
Polycrate API 0.11.21 addresses two critical issues: S3 buckets with Ceph backend now display …
With Polycrate CLI 0.29.15, we have resolved the root cause of a persistent bug where endpoints with …