v3.5.0

• feature: folder support. This release will put every dashboard in a folder named after the origin namespace of the dashboard. Thanks @HubertStefanski
• feature: requests and limits now configurable for init containers too. Thanks @stgarf
• feature: image pull secrets can now be specified for service accounts. Thanks @AdheipSingh
• feature: Stackdriver datasources are now supported. See #225
• fix: preferService was ignored when ingress or route were not disabled. Thanks @HubertStefanski

NOTE: when updating to this release, dashboards will be put into folders. The folder name is the namespace of the dashboard. We plan to extend folder support and make it more configurable in future releases.

v3.4.0

• feature: persistent storage for the Grafana Data directory. Thanks @polefishu
• feature: namespace selector when --scan-all is used. Thanks @R-Lawton
• feature: configuring Grafana container security context. Thanks @AdheipSingh
• fix: deadlock caused by readiness probe when upgrading. Thanks @Rajagopalan-Ranganathan
• cleanup remove useless function call. Thanks @kqzh

v3.3.0

• email_attribute_path now configurable in oauth.generic . Thanks @mjattiot
• Fixed a data race condition when importing from a large number of namespaces. Thanks @obrienrobert
• Termination grace period is now configurable in the deployment section. Thanks @AdheipSingh
• Fixed an issue where dashboards would not get reimported when operator and Grafana are restarted. Thanks @renanqts

v3.2.0

• Use crypto/rand instead of math/rand
• Allow additional fields for data sources jsonData and secureJsonData. See #157 , thanks @Slach
• ConfigMaps as dashboard targets. See #139 , thanks @phillebaba

v3.1.0

• Feature: support for Tolerations, Affinity, SecurityContext and NodeSelector in the deployment config. Thanks @AdheipSingh
• Fix: allow setting role_attribute_path in auth.generic_oauth. Thanks @renanqts
• Feature: Includes ansible scripts for installing and configuring the operator. Thanks @itewk
• Feature: Gitlab supported in auth configuration. Thanks @shivpathak
• Fix: does no longer overwrite preexisting annotations.
• Fix: missing permissions for events. Thanks @Slach
• Fix: use Ingress hostname from CR if available to fix #136 and #137
• Fix: do not reuse dashboard IDs to prevent dashboards overwriting other dashboards in certain cases. Thanks @davidkirwan @R-Lawton

v3.0.2

Following advice from here: https://golang.org/cmd/go/#hdr-Module_compatibility_and_semantic_versioning

We've determined that we need to make the change to the module version:

git diff master go_mod_version_bamp 
diff --git a/go.mod b/go.mod
index 2460be6..620de8f 100644
--- a/go.mod
+++ b/go.mod
@@ -1,4 +1,4 @@
-module github.com/integr8ly/grafana-operator
+module github.com/integr8ly/grafana-operator/v3
 
 go 1.13
 

Users of the operator will now need to do the following to access the latest version:

• need to now import grafana-operator eg: go get github.com/integr8ly/grafana-operator/[email protected]
• need to modify imports eg:

import (
	grafanav1alpha1 "github.com/integr8ly/grafana-operator/v3/pkg/apis/integreatly/v1alpha1"
...

v3.0.1

• fixes #88 (volume mounts for additional configmaps incorrectly prefixed)
• adds an example for #63 and fixes a bug where invalid configuration for remote cache was generated

v3.0.0

• Defaults to Grafana 6.5.1
• Dashboards are now imported using the Grafana API directly, no more ConfigMap. Thanks @lleszczu for the initial implementation
• Dashboards and Datasources no longer have finalizers. They don't require the Operator to be present to delete them anymore.
• Update reconciliation strategy, many properties of the created resources are now configurable including the resources requirements for the Grafana Pod. Thanks @AndreasBergmeier6176 for pointing this out and providing an implementation.
• Updated to operator-sdk v0.12.0.
• Uses Go modules instead of dep now.

Notes

see the changelog for more info about upgrading.

v2.0.0

• #52: New flag --namespaces to watch for dashboards in a list of namespaces.
• #49: Secrets and ConfigMaps can be mounted into the Grafana Pod as Volumes.
• #43: Dashboards can be imported from URL.
• #51: Many customization options for Service, Ingress and Route (contributed by @gambol99)
• #51: Supports scaling by setting the number of replicas in the Grafana CR (contributed by @gambol99)

Removed in this version:

• --openshift: The operator auto detects the environment now and creates a Route when on OpenShift.

Changed in this version:

• The operator no longer creates an Ingress or Route automatically. It has to be enabled explicitely by setting spec.ingress.enabled in the Grafana CR.

v1.4.0

#40 (fix: adding jsonData and secureJsonData to data sources only accepted strings)
#39 (feat: custom value for the app label by means of an operator flag)