You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Please add support to specify default leader election Lease lock name or generate the name to something unique which includes instance-id.
Use Cases
I use Argo Rollouts in a distributed setup where multiple Argo Rollouts Controllers with --instance-id specified are connecting to a central cluster where a central Argo Rollouts Controller is running. The distributed controllers are using kubeconfig of the central cluster, so all of them are are detecting the same Lease (which is currently specified in the code as constant valueargo-rollouts-controller-lock and therefore only one is elected, even though they are differentiated because of --instance-id flag.
According to Kubernetes documentation, best practice in case multiple instances are deployed of a component is like using a hashed value or a prefix to avoid collisions for the Leases:
If a cluster operator or another end user could deploy multiple instances of a component, select a name prefix and pick a mechanism (such as hash of the name of the Deployment) to avoid name collisions for the Leases.
You can use another approach so long as it achieves the same outcome: different software products do not conflict with one another.
Message from the maintainers:
Impacted by this bug? Give it a 👍. We prioritize the issues with the most 👍.
The text was updated successfully, but these errors were encountered:
Summary
Please add support to specify default leader election
Lease
lock name or generate the name to something unique which includes instance-id.Use Cases
I use Argo Rollouts in a distributed setup where multiple Argo Rollouts Controllers with
--instance-id
specified are connecting to a central cluster where a central Argo Rollouts Controller is running. The distributed controllers are using kubeconfig of the central cluster, so all of them are are detecting the same Lease (which is currently specified in the code as constant valueargo-rollouts-controller-lock
and therefore only one is elected, even though they are differentiated because of --instance-id flag.According to Kubernetes documentation, best practice in case multiple instances are deployed of a component is like using a hashed value or a prefix to avoid collisions for the Leases:
Message from the maintainers:
Impacted by this bug? Give it a 👍. We prioritize the issues with the most 👍.
The text was updated successfully, but these errors were encountered: