External Redis config

[j771]

Is there a way to setup Spinnaker deployed with the Armory operator to use an external redis?
I see there is some documentation here but I tried this and it did not work: https://docs.armory.io/docs/installation/operator-reference/deploy/
Maybe set the configs from the above link and also set the service-settings/redis.yml as stated here: https://spinnaker.io/setup/productionize/caching/externalize-redis/#configure-a-spinnaker-wide-redis

Not sure how to setup the redis.yml for Armory deploy.
Would it be something like the following?

spec:
  spinnakerConfig:
    service-settings:
      redis:
        redis.yml : |
          overrideBaseUrl: redis://admin:[email protected]:6379
          skipLifeCycleManagement: true

    profiles:
      gate:
        redis:
          configuration:
            secure: true

And for deployment environment config?

spec:
  spinnakerConfig:
    config:
      deploymentEnvironment:
        haServices:
           clouddriver:
             enabled: true
             disableClouddriverRoDeck: redis://admin:[email protected]:6379
             redisMasterEndpoint: redis://admin:[email protected]:6379
             redisSlaveEndpoint: redis://admin:[email protected]:6379
             redisSlaveDeckEndpoint: redis://admin:[email protected]:6379
           echo:
             enabled: true

[devorbitus]

Here is an example of using an external redis https://github.com/armory/spinnaker-kustomize-patches/blob/master/persistence/patch-redis.yml

[j771]

Thanks, I found that a few days ago and I was able to set it up.
I found that I was able to use a k8s secret for the redis url, which is good.

spec:
  spinnakerConfig:
    profiles:
      gate:
        redis:
          configuration:
            secure: true

    service-settings:
      redis:
        overrideBaseUrl: encrypted:k8s!n:<SECRET_NAME>!k:<SECRET_KEY>
        skipLifeCycleManagement: true

Although the Armory operator seems to be stuck in Updating mode for 12 hours and has not deleted the old redis deployed by the operator before switching to the external redis-ha deploy.

I am seeing this in the operator logs:

{"level":"error","ts":1610705811.6199894,"logger":"controller-runtime.controller","msg":"Reconciler error","controller":"spinnakerservice-controller","request":"spinnaker-operator/spinnaker","error":"Operation cannot be fulfilled on spinnakerservices.spinnaker.io "spinnaker": the object has been modified; please apply your changes to the latest version and try again","stacktrace":"github.com/go-logr/zapr.(*zapLogger).Error\n\t/opt/spinnaker-operator/build/vendor/github.com/go-logr/zapr/zapr.go:128\nsigs.k8s.io/controller-runtime/pkg/internal/controller.(*Controller).reconcileHandler\n\t/opt/spinnaker-operator/build/vendor/sigs.k8s.io/controller-runtime/pkg/internal/controller/controller.go:218\nsigs.k8s.io/controller-runtime/pkg/internal/controller.(*Controller).processNextWorkItem\n\t/opt/spinnaker-operator/build/vendor/sigs.k8s.io/controller-runtime/pkg/internal/controller/controller.go:192\nsigs.k8s.io/controller-runtime/pkg/internal/controller.(*Controller).worker\n\t/opt/spinnaker-operator/build/vendor/sigs.k8s.io/controller-runtime/pkg/internal/controller/controller.go:171\nk8s.io/apimachinery/pkg/util/wait.JitterUntil.func1\n\t/opt/spinnaker-operator/build/vendor/k8s.io/apimachinery/pkg/util/wait/wait.go:152\nk8s.io/apimachinery/pkg/util/wait.JitterUntil\n\t/opt/spinnaker-operator/build/vendor/k8s.io/apimachinery/pkg/util/wait/wait.go:153\nk8s.io/apimachinery/pkg/util/wait.Until\n\t/opt/spinnaker-operator/build/vendor/k8s.io/apimachinery/pkg/util/wait/wait.go:88"}

[devorbitus]

Een regular Halyard based deployments would not remove the previously deployed redis pods.

Note: By setting skipLifeCycleManagement we are telling Halyard to stop deploying/check the status of the Redis instance. If Halyard has already created a Redis instance, you will have to manually delete it. -> https://spinnaker.io/setup/productionize/caching/externalize-redis/

Operator is currently running Halyard under the hood so it would be the same.

[j771]

Thanks, I thought that maybe the operator would have taken care of that. I will manually delete the old redis, thank you.

[j771]

I manually deleted the old redis but the spinsvc is still in a Status Updating.

[dogonthehorizon]

@j771 are you still seeing issues with updating the Operator? There were some changes in 1.2.x around how status was being reported that may resolve the behavior you were seeing.

[j771]

I have updated to the latest version of the operator and when I did the update I did not run into any issues.