investigate what happens when a machine with a duplicate name of an existing machine is created in a different namespace

[chuckha]

/kind bug

I think this is a bug and this issue is mostly here to remind me to look into this at some point and flesh out the actual bug details or close if it's not an issue.

When two machines share a name but not a namespace, does everything still work as expected?

[detiber]

I don't expect there to be issues as long as they do not also share the same "cluster" name. That said, we currently do not prevent multiple "clusters" existing in different namespaces, however that would definitely cause issues today.

[ashish-amarnath]

This will be a problem based on how we, currently, look if a machine exists. Specifically, when we look for instances by tags. https://github.com/kubernetes-sigs/cluster-api-provider-aws/blob/release-0.3/pkg/cloud/aws/services/ec2/instances.go#L66
Even more specifically, that problem is
https://github.com/kubernetes-sigs/cluster-api-provider-aws/blob/release-0.3/pkg/cloud/aws/services/ec2/instances.go#L86-L88
This will not be an issue if the machines belong to a different cluster (and therefore a different VPC?).

[ncdc]

I'm curious why we need to query for an instance by tags if we have instance ID?

[ashish-amarnath]

We query by tags only when we don't find the instanceID
https://github.com/kubernetes-sigs/cluster-api-provider-aws/blob/release-0.3/pkg/cloud/aws/services/ec2/instances.go#L481

But when there is a naming conflict, I think, we'll end up w/ 2 machine objects being liked to the same instance ID.

[ncdc]

Right, but why? 😄

IMHO, if a Machine does not have spec.providerID, we must do a create operation. If it does have spec.providerID, then we should be doing existence/health checks and setting the error state appropriately if the VM is not pending or running (+ maybe shutting down & stopped too).

[detiber]

@ncdc The reason for the check is in case we create the instance but we are not able to update the Machine object for some reason.

[ncdc]

Ok, that's fair.

[fejta-bot]

Issues go stale after 90d of inactivity.
Mark the issue as fresh with /remove-lifecycle stale.
Stale issues rot after an additional 30d of inactivity and eventually close.

If this issue is safe to close now please do so with /close.

Send feedback to sig-testing, kubernetes/test-infra and/or fejta.
/lifecycle stale

[fejta-bot]

Stale issues rot after 30d of inactivity.
Mark the issue as fresh with /remove-lifecycle rotten.
Rotten issues close after an additional 30d of inactivity.

If this issue is safe to close now please do so with /close.

Send feedback to sig-testing, kubernetes/test-infra and/or fejta.
/lifecycle rotten

[vincepri]

/lifecycle frozen

[randomvariable]

/help

[k8s-ci-robot]

@randomvariable:
This request has been marked as needing help from a contributor.

Please ensure the request meets the requirements listed here.

If this request no longer meets these requirements, the label can be removed
by commenting with the /remove-help command.

In response to this:

/help

Instructions for interacting with me using PR comments are available here. If you have questions or suggestions related to my behavior, please file an issue against the kubernetes/test-infra repository.

[MarcusNoble]

#969 seems to answer this question, is this issue still relevant?

[randomvariable]

thanks @MarcusNoble .

/close
since that one supersedes.

Going to take to SIG Cloud Provider for further discussion.

[k8s-ci-robot]

@randomvariable: Closing this issue.

In response to this:

thanks @MarcusNoble .

/close
since that one supersedes.

Going to take to SIG Cloud Provider for further discussion.

Instructions for interacting with me using PR comments are available here. If you have questions or suggestions related to my behavior, please file an issue against the kubernetes/test-infra repository.