Skip to content

Latest commit

 

History

History
451 lines (273 loc) · 12.2 KB

CHANGELOG.md

File metadata and controls

451 lines (273 loc) · 12.2 KB
Raw

Changelog

All notable changes to this project will be documented in this file.

The format is based on Keep a Changelog, and this project adheres to Semantic Versioning.

Unreleased

0.16.0 - 2024-11-26

Added

  • Add kubernetes events logging in Alloy.
  • Add support for Private CAs in alloy logs.
  • Add KubeEventsLogger option and related methods in loggedCLuster package.
  • Add events-logger flag in the operator.
  • Add toggle for events-logger in observability-bundle configmap.
  • Add tests for alloy-events in events-logger-config.

Changed

  • Disable grafana agent usage data reporting.
  • Move Grafana-Agent config to a template instead of go structs.

Fixed

  • Fix logging-config unit tests

0.15.2 - 2024-11-13

Added

  • Enable VPA on alloy if the deployed alloy version supports it.

0.15.1 - 2024-11-04

Fixed

  • Fixes the current version that enabled alloy-logs as the new secret mechanism only works with alloy 0.4.0 which is is the observability bundle 1.6.0

0.15.0 - 2024-10-31

Changed

  • Reconcile clusters when the observability bundle version changes.

Fixed

  • Disable crd installation from grafana agent as this is causing issues with the new v29 releases.

0.14.0 - 2024-10-29

Changed

  • Change default logging agent to Alloy instead of Promtail.

0.13.0 - 2024-10-24

Added

  • Add manual e2e testing procedure and script.
  • [Alloy] Add capability to dynamically configure log targets using PodLogs #3618
    • Changes Alloy config to use PodLogs for Kubernetes pods discovery.
    • There is a performance impact on Kubernetes API
    • Available from observability-bundle v17.0.0

Changed

  • [Alloy] Enable clustering
  • Expose healthcheck port for kube-linter.

Removed

  • Remove the alloy-secret resource which is no longer needed as Alloy secret was moved into logging-secret

Fixed

  • Fix circleci config.

0.12.1 - 2024-09-23

Fixed

  • Fix usage of structured metadata for clusters before v20.

0.12.0 - 2024-09-23

Changed

  • Move high cardinality values into structured-metadata:
    • kubernetes audit log resource label
    • filename label
    • log output stream label
  • Rename node_name label into node to match metric label.

0.11.2 - 2024-09-18

Fixed

  • Fix v0.11.1 release was not published

0.11.1 - 2024-09-18

Fixed

  • Fix v0.11.0 release was not published

0.11.0 - 2024-09-18

Fixed

  • Fix Alloy logs secret handling

0.10.0 - 2024-09-10

Changed

  • Use grafana multi-tenant-proxy public types instead of vendoring them.

Fixed

  • Disable logger development mode to avoid panicking, use zap as logger.

Removed

  • Delete old loki-auth-proxy configuration in favor of the new grafana-auth-proxy.
  • Remove old loki-auth reconciler.

0.9.0 - 2024-09-03

Changed

  • Change the datasource url to be the multi-tenant-proxy in front of the loki-gateway.
  • Add secret management for the proxy by duplicating loki-auth.

Fixed

  • Fix incorrect alloy security context.

0.8.0 - 2024-08-27

Added

  • Add helm chart templating test in ci pipeline.
  • Add tests with ats in ci pipeline.

0.7.3 - 2024-08-13

Fixed

  • Fix required observability bundle version to run Alloy as logging agent.

0.7.2 - 2024-08-12

Fixed

  • Fix alloy-secret naming
    • Rename secret to alloy-logging-secret
    • Prefix secret with cluster name
    • Create secret in the cluster namespace

0.7.1 - 2024-08-08

Fixed

  • Fix incorrect Alloy secret and Alloy config templating.
  • Rename alloy-logs to came case alloyLogs in the observability-bundle config.

0.7.0 - 2024-07-19

Added

  • Add support for Alloy as logging agent Add --logging-agent flag to toggle between Promtail and Alloy

0.6.0 - 2024-07-09

Changed

  • Use a deployment for the grafana-agent instance used to collect kubernetes events to avoid using too much resources on clusters as long as we use promtail.

0.5.5 - 2024-06-14

Fixed

  • Fix reconciliation errors when adding or removing the finalizer on the Cluster CR.

0.5.4 - 2024-04-23

Changed

  • Replace promtail occurrences by logging to be more generic.

Fixed

  • Delete leftover configmaps and secrets while cluster deleting.

0.5.3 - 2024-04-09

Changed

  • Reduce audit log cardinality by ignoring rotated audit log files to avoid duplicate audit logs.

0.5.2 - 2024-03-06

Changed

  • Update deprecated targetPort to port in PodMonitor.

0.5.1 - 2024-02-21

Fixed

  • This feature fixes reconciliation by support requeuing of failed request that do not necessarily need to be an error (missing app due to slow cluster boostrapping).

0.5.0 - 2024-02-19

Removed

  • Remove multi-tenant proxy restart hack.

0.4.4 - 2024-01-22

Changed

  • Push to CAPV.

0.4.3 - 2024-01-17

Fixed

  • Ignore not found error for clusters that have logging disabled.

0.4.2 - 2024-01-11

Fixed

  • Watch observability-bundle apps to handle bundle migration to v1.0.0.

0.4.1 - 2024-01-09

Fixed

  • Fix reconcile errors (grafana-agent-config, promtail-wiring, cluster not found).

0.4.0 - 2024-01-04

Added

  • Expose profiles in the controller and add conditional profiling annotations in the deployment.

Changed

  • Configure gsoci.azurecr.io as the default container image registry.
  • Drop system audit logs from Promtail's scrape target

Fixed

  • Replace systemd_unit label with syslog identifier for system logs without systemd_unit label
  • Fixed podmonitor

0.3.1 - 2023-12-04

Changed

  • enable logging on WCs by default.
  • push to CAPZ and CAPVCD collections

0.3.0 - 2023-11-21

Changed

  • Add labels to kubernetes audit logs to reduce rate limiting and help discovering logs.

0.2.2 - 2023-11-14

Fixed

  • Fix grafana-agent configMap creation on CAPI.

0.2.1 - 2023-11-10

Changed

  • default logging behavior on WCs reverted to disable.

0.2.0 - 2023-11-09

Added

  • Configure grafana-agent config to grab Kubernetes Events and send them to Loki.
  • Create grafana-agent extra secret to store logging write credentials.
  • Prepare some stuff for CAPI.
  • Upgrade go dependencies.

Changed

  • Change default logging behavior on WCs >= 19.1.0. Logging is now enabled by default.
  • Improve network policy and minor go fixes.

0.1.4 - 2023-10-31

Changed

  • Push to CAPA app collection.

0.1.3 - 2023-10-18

Changed

  • Configure correct app depending on observability-bundle version.

0.1.2 - 2023-10-17

Fixed

  • Revert support for observability-bundle 1.0.0.

0.1.1 - 2023-10-17

Added

  • Add support for observability-bundle 1.0.0.

0.1.0 - 2023-10-17

Changed

  • Only workload clusters release >= v19.1.0 can enable logging.
  • each cluster has a dedicated user
  • each cluster sends data as a different tenant
  • update logging-credentials secret format

0.0.7 - 2023-10-03

Added

  • Audit logs in promtail config.
  • Add condition for PSP installation in helm chart.

Changed

  • Logs labels updated to ease navigation.

0.0.6 - 2023-09-28

Fixed

  • Ensure we do not delete observability-bundle user configs for workload clusters.

0.0.5 - 2023-09-20

Added

  • Scrape logs from kube-system and giantswarm namespaces only for WC clusters.

0.0.4 - 2023-09-18

Changed

  • Adapted code to handle promtail deployment in WCs.

0.0.3 - 2023-07-27

Fixed

  • Add missing RBAC access to apps/deployment resources.

0.0.2 - 2023-07-26

Added

  • promtail-config reconciler: creates promtail-config as extra-values.

Changed

  • Push app to aws-app-catalog
  • Commented reconcilers creation for Vintage WC and CAPI clusters - not supported yet.
  • image tag is defined from chart version

Fixed

  • PSP permissions to update app
  • fix CODEOWNERS

0.0.1 - 2023-07-13

Added

  • Add Helm chart
  • Implement LoggingReconciler abstraction
  • Implement promtrail-wiring resource
  • Implement promtail-toggle resource
  • Implement finalizer handling
  • Add reconciler.Interface
  • Add controller for Vintage Management Cluster via corev1.Service
  • Add controller for Cluster API, cluster.x-k8s.io/v1beta1
  • Add operator basics with kubebuilder
  • Add '--vintage' toggle
  • Add controller for Workload Management Cluster using cluster.x-k8s.io/v1beta1```