v3.0.3

What's Changed

• Hotfix: Fix issue where the intents operator would not be able to restart due to circular dependency on its webhook server, introduced in the previous version. Fix by @omris94 in #567
  Note: When you upgrade, a new Pod will spin up, the intents-operator-webhook-server, to resolve this issue. If you have specified configuration for the intents-operator Pods, you can safely duplicate the values from Values.intentsOperator.operator to Values.intentsOperator.webhookServer to achieve the same effect.

Full Changelog: v3.0.2...v3.0.3

v3.0.2

What's Changed

• Fix a bug where errors were reported on IAM ClientIntents update or deletion, by detecting that the cloud role was not found, and re-queuing or completing the request by @otterobert in #555
• Fix intents-operator sending STARTED telemetry with 0-count, which is inconsistent with other operators by @amitlicht in #556
• Fixed bug where the AllowExternalTraffic configuration was not set to its default value by @zohar7ch in #558
• When specifying Internet intents with network policy egress enforcement disabled, the operator now adds a Warning to the Events on the ClientIntents resource by @zohar7ch in #561
• Added support for reusing the webhook TLS certificate across different deployments to prevent cases where operator takes too long to start up by @orishoshan in #560

Full Changelog: v3.0.1...v3.0.2

v3.0.1

What's Changed

• Periodically Check and cleanup orphaned resources in azure by @otterobert in #553
• Fix bug where operator config report would report incorrect Istio/Linkerd installation status upon startup by @zohar7ch in #554

New Contributors

• @zohar7ch made their first contribution in #554

Full Changelog: v3.0.0...v3.0.1

v3.0.0

What's Changed

• Support new API version for ClientIntents: v2beta1 by @omris94 in #552

We have revamped the format for ClientIntents to make them easier to understand and make it possible to generate more restrictive policies by default. Read more on the docs >>

The two primary changes is that the word service is no longer used, except to mean a Kubernetes Service; before, it could mean an Otterize Service or a Kubernetes Service, which was confusing. Instead, we now use workload . calls have also been renamed targets , and many smaller changes to the structure to improve

What happens to your existing ClientIntents? Don’t worry, the change is backwards compatible, and nothing changes unless you explicitly upgrade. If you’re a customer, we’ll reach out to explain and plan together. If you’re using the open source, upgrading to the next major version of the otterize-kubernetes Helm chart, vX.X.X will make ClientIntents v2 the default. You can still continue applying ClientIntents with apiVersion v1alpha3, and they will be converted by the Intents Operator to v2beta1.

Full Changelog: v2.0.42...v3.0.0

v2.0.42

What's Changed

• Solve a potential race on AWS IAM integration where policy deletion would error because the related IAM role was not found by @otterobert in #551

Full Changelog: v2.0.41...v2.0.42

v2.0.41

What's Changed

• Fix ConcurrentModification errors for aws policies by @otterobert in #550
• Report linkerd configuration to Otterize Cloud by @evyatarmeged in #547

Full Changelog: v2.0.40...v2.0.41

v2.0.40

What's Changed

• Fixed possible panic caused by calling wrong event recorder when handling effective policy for Linkerd by @evyatarmeged in #548
• Fix bug where AWS IAM roles could be created for a particular workload even when disabled for that particular workload by @amitlicht in #549

Full Changelog: v2.0.39...v2.0.40

v2.0.39

What's Changed

• Update terminology: Rename annotation intents.otterize.com/service-name to intents.otterize.com/workload-name, retaining support for backward compatibility by @omris94 in #542
• Fix Azure API version error when applying Azure ClientIntents on KeyVault scopes, by querying for supported API versions by resource type by @amitlicht in #546

Full Changelog: v2.0.38...v2.0.39

v2.0.38

What's Changed

• Fixing no authrization error for PostgreSQL in managed databases by @evyatarmeged in #539
• Fixed bug where pod label updates fail on short-lived containers by @evyatarmeged in #541
• Rename spec.service to spec.workload in KafkaServerConfig API version v2alpha1 by @omris94 in #540
• Align linkerd RBAC permissions for the intents-operator cluster role with the helm-chart by @omris94 in #544
• Support reporting applied ClientIntents targeting Kubernetes services to the cloud by @omris94 in #545

Full Changelog: v2.0.37...v2.0.38

v2.0.37

What's Changed

• Improve CPU consumption by skipping populateReferencedKubernetesServices logic when egress enforcement is disabled by @omris94 in #538

Full Changelog: v2.0.36...v2.0.37