CHANGELOG.md

Changelog

Future

v0.10

v0.10.2

Bug fixes

• Fixed a bug where the proxy setting was not properly passed when using immutable images (#394)
• Fixed a bug where the proxy value was not properly used in the Operator (#396)

Other changes

• Removed PodSecurityPolicies since they got removed with Kubernetes 1.22 (#392)
• Updated the apiVersion of the CRD from v1beta1 to v1 since v1beta1 got removed with Kubernetes 1.22 (#393)
• Added a separate openshift3.11.yaml as release artifact to allow compatibility with deprecated apiVersions for the CRD (#395)

v0.10.1

Other changes

• Updated the Operator image to rely on the latest ubi-minimal base image (#389)

v0.10.0

Bug fixes

• Don't look at the cluster version when deploying the OneAgent using immutable images. Under certain conditions this may stop the Operator from deploying the OneAgent at all (#376)
• Upgrade OneAgent Pods using the immutable image by looking at the version label embedded on the images (#376)

Other changes

• Upgrade to Operator SDK 1.3 (#351)
  • Use ConfigMaps for leases (#367)
• Fix version and deployment instructions for dev builds (#379)

v0.9

v0.9.5

Bug fixes

• Adapted the update interval to only do the request very 15 minutes (#368)

v0.9.4

Bug fixes

• Temporary files for the injection via webhook now get placed on an emptyDir to support readonlyFileSystems (#353)
• ImagePullPolicy for install-oneagent container used by OneAgentAPM got changed to Always to allow updates (#358)
• Nodes in a non-working state will not get added as empty entries to the init.sh script anymore (#359)
• Querying for the Dynatrace cluster version will just be done per OneAgent resource when using the immutable image (#360)

v0.9.3

• Minor release just for Helm

v0.9.2

Features

• Added property to change default C standard library (#341)

Bug fixes

• Don't restart OneAgent Pods if OneAgent version has been downgraded and those Pods were already restarted (#339, #345)

Other changes

• Updated the UBI-minimal base image from version 8.2 to 8.3 (#344)

v0.9.1

• Minor release just for OperatorHub

v0.9.0

Features

• Provide Prometheus metrics for the Operator pod on port 8080, and Webhook Pod on ports 8383 and 8484 (#305)

• Control whether the init container crashes in case of download failures through the oneagent.dynatrace.com/failure-policy: fail Pod annotation, off by default (#288)

  • Regardless of the annotation, if the unzip operation fails, a file package.zip will be included on the target directory for debugging purposes.

• Resource limits and requests for the OneAgentAPM initContainer are configurable on the .spec.resources field (#332)

• Early Adopter: support full-stack OneAgent running on unprivileged mode (#324, #333)

Bug fixes

• Logged errors when API token is missing on OneAgentAPM's secret (#298)
• Fixed printing the name of the used token secret for OneAgent instances (311)
• Fixed setting instances metadata when auto-update is disabled (#313)
• Fixed logging problem - incorrect stackdriver severity on GCP (#318)

Other changes

• Added support for immutable OneAgent images - waiting for support on Dynatrace cluster (#300, #286, #290, #301)
• Added check if cluster and agent versions are compatible with immutable images (#314, #334)
• Immutable image mode is disabled when a custom installer URL annotation is set, or .spec.useImmutableImage is false (#306)
• Pod and node metadata added for the OneAgent (#294, #295, #308, #325, #326)
• Code cleanup to remove unused functions, variables and beautify the code (#302)
• Sped up TravisCI duration (#310, #312)
• Upgrade to Go 1.15 (#310)
• Add linter to TravisCI pipeline (#316)
• App-only init container will log an warning when the full-stack OneAgent has been injected on it (#323)
• Improve error message when OneAgentAPM is missing (#327)
• Improve descriptions on cr.yaml example (#328)

v0.8

v0.8.2

Bug fixes

• Reworked update mechanism to prevent downgrades (#320)

v0.8.1

Features

• Publish Operator stable images also to Quay (#304)

Bug fixes

• Update status of OneAgentAPM if token is missing (#285, #287)
• Marked for termination events are now sent when a node is deleted, or when it's cordoned, and then periodically after each hour while in that state (#279, #303)
• Operator was updating the DaemonSet even if there were no changes (#289)
• Certificates secret not updated on renewal, causing renewals every 5 minutes (#297)
• Add OLM manifests for v0.8.1 (#307)

v0.8.0

Features

• Implement webhook to inject the OneAgent in App-only mode (#234, #237, #239, #250)
  • This feature can be enabled by setting the label oneagent.dynatrace.com/instance: <oneagent-object-name> on the namespaces to monitor.
  • CA and server certificates are generated for the webhook by the Operator, and renewed automatically after 365 and 7 days, respectively (#244)
  • OneAgent app-only package and logs will be stored on /opt/dynatrace/oneagent-paas inside the containers by default. It can be configured with the oneagent.dynatrace.com/install-path annotation on Pods (#251)
  • OneAgent app-only package will be downloaded from the provided tenant by default. It can be configured with the oneagent.dynatrace.com/installer-url annotation on Pods (#258)
  • Certificates location can be configured on the webhook server with the --certs-dir, --cert, and --cert-key command line arguments (#261)
  • When setting a custom installer url the authentication header won't be sent (#264)
• Added a setting to configure a NetworkZone via the CR (#270)

Bug fixes

• Phase now gets set to 'Deploying' while the OneAgent gets updated (#267)

Other changes

• Removed kubernetes.yaml and openshift.yaml from master and generate them with kustomize instead (#238, #254)
• Updated the Go version from 1.13 to 1.14 (#242)
• Updated the Operator SDK version from 0.15.0 to 0.17.0 (#243)
• The different operator and webhook modes are encapsulated in a single binary (#252, #253)
• Webhook's init container only downloads 64bits package (#256)
• Include Service and MutatingWebhookConfiguration objects in manifests (#262, #266)
• Upgrade base image to ubi-minimal:8.2 (#255)
• Include Operator version as a custom property for hosts (#212)
• Ignore hosts that haven't seen in the last 30 minutes when looking for hosts (#271, #257)
• Adjust permissions for the webhook (#263)
• Refactor workflow from OneAgent controller (#268)
• Automatically update conditions if migrating from earlier Operator versions (#269)
• Remove unused metadata from webhook-injected Pods (#272)
• Changes in preparation for v0.8.0 release (#273, #274)
• Add OLM manifests for v0.8.0 (#275)

v0.7

v0.7.1

Bug fixes

• Marked for Termination events are not a point in time instead of a time range of a few minutes (#229)
• Fixed error message when OneAgent has been already removed from the cache but the node was still there (#232)

Other changes

• Added environment variable 'RELATED_IMAGE_DYNATRACE_ONEAGENT' as preparation for RedHat marketplace release (#228)
• Fixed some problems with the current Travis CI build (#230)

v0.7.0

Breaking changes

• This version drops support for Kubernetes 1.11, 1.12, and 1.13 (#219, #220)

Features

• Separated the logic for watching the nodes into nodes_controller to handle scaling correctly (#189, #196)
• Show operator phase in the status.phase field of the OneAgent object (#197)
• Build ARM64 images for the Operator (#201, #211)
• No longer change the OneAgent .spec section to set defaults (#206)
• Added a setting to configure a proxy via the CR (#207)
• Added a setting to add custom CA certificates via the CR - These changes are only done for the Operator image as of now and the changes in the OneAgent image are in progress (#208)
• Added proper error handling for Dynatrace API quota limit (#216)

Bug fixes

• Handle sporadic (and benign) race conditions where the error below would appear (#194),

  Operation cannot be fulfilled on oneagents.dynatrace.com \"oneagent\": the object has been modified; please apply your changes to the latest version and try again
  

• Proxy environment variables (e.g., http_proxy, etc.) can be ignored on Operator container when skipCertCheck is true (#204)
• Istio objects don't have an owner object, so wouldn't get removed if the OneAgent object is deleted (#217)

Other changes

• As part of the support for ARM (#201, #203)
  • Migrate CI/CD workflow from CircleCI to TravisCI
  • Development snapshot images are now being published to Docker Hub
• Support deprecation of beta.kubernetes.io/arch and beta.kubernetes.io/os labels (#199)
• Update to Operator SDK 0.15.1 (#200)
• Initial work to ease release automation (#198)
• Added automatic creation of CSV file for OLM (#210)
• Now Marked for Termination events will be sent only for deleted Nodes (#213, #214, #223)
• Use v1 instead of v1beta1 for rbac.authorization.k8s.io objects (#215)
• Add OLM manifests for v0.7.0 release (#226)

v0.6

v0.6.0

Features

• Improve error logging from Dynatrace API requests on Operator (#185)
• Allow custom DNS Policy for OneAgent pods (#162)
• Add OpenAPI V3 Schema to CRD objects (#171)
• Operator log entries now use ISO-8601 timestamps (e.g., "2019-10-30T12:59:43.717+0100") (#159)
• The service account for pods can now be customized (#182, #187)
• Custom labels can be added to pods (#183, #191)
• Validate tokens for OneAgent and show results as conditions on OneAgent status section (#188, #190)

Bug fixes

• Operator needs to be restarted after Istio is installed. Fixed on controller-runtime v0.3.0 (#172, controller-runtime#554)

Other changes

• Installation steps on Readme are now for stable releases (#205)
• Most operations now use HTTP Header for authentication with Dynatrace API (#167)
• Operator Docker images have been merged, and are now based on UBI (#179)
• Update to nested OLM bundle structure (#163)
• Code style improvements (#158, #175)
• Update to Operator SDK 0.12.0 and Go modules (#157, #172)
• Using istio.io/client-go to manage Istio objects (#174)
• Add OLM manifests for v0.6.0 (#193)

v0.5

v0.5.4

• Service update to get the latest version of the base image for RedHat Container Catalogue where some vulnerabilities have been fixed.

v0.5.3

• Allow to customize the service account used for creating the OneAgent pods in preparation to release on GKE marketplace (#182, #187)

v0.5.2

• Fixes: non-default Secret for Dynatrace tokens ignored (#168)

v0.5.1

• Fixes: panics when handling node notifications for unmatching OneAgent node selectors (#160, #161)

v0.5.0

• Better detection and handling of node scaling events. (#153, #154)
• Improved documentation for installation of the Operator in OCP 3.11 environments. (#156)

Older versions

v0.4.2

• Bug fixes

v0.4.1

• Bug fixes

v0.4.0

• Support automatic configuration of Istio to allow communication from the OneAgent pods to the Dynatrace environment.
• Added support for Kubernetes 1.16

v0.3.1

• Bug fixes

v0.3.0

• Add configuration field on OneAgent CR to set OneAgent pods' Priority Class.
• Requires Kubernetes 1.11+/OCP 3.11+

v0.2.1

• Add configuration field on OneAgent CR to disable automatic updates from the agent.
• This is the last version supporting Kubernetes 1.9 and 1.10, as well as OCP 3.9 and 3.10.

v0.2.0

• Add configuration field on OneAgent Custom Resource for Node Selector and Resource Quotas of OneAgent pods.

v0.1.0

• Initial release.
• Support for Kubernetes 1.9 and OpenShift Container Platform 3.9 or higher.