Spotinst: Avoid unnecessary duplication of tasks

[liranp]

This PR fixes the regression introduced by #9699 (reported by @povils): renaming the instance group results in nodeup failing to start with the following errors:

$ journalctl --no-pager -o cat -u kops-configuration
Starting Run kops bootstrap (nodeup)...
nodeup version 1.20.0-alpha.1
I0121 17:47:52.707955    1846 s3context.go:213] found bucket in region "us-west-2"
I0121 17:47:52.708041    1846 s3fs.go:290] Reading file "s3://<CONFIG_BASE>/cluster.spec"
I0121 17:47:52.742606    1846 s3fs.go:290] Reading file "s3://<CONFIG_BASE>/instancegroup/nodes.<CLUSTER_NAME>"
W0121 17:47:52.750671    1846 main.go:133] got error running nodeup (will retry in 30s): error loading InstanceGroup "s3://<CONFIG_BASE>/instancegroup/nodes.<CLUSTER_NAME>": file does not exist

[hakman]

@liranp Can you edit the commit message?

[liranp]

Done it already. Sorry about that.

[povils]

Thanks for picking this up! will it go to 1.19 ?

[hakman]

No worries, thanks!
/lgtm
/approve

[k8s-ci-robot]

[APPROVALNOTIFIER] This PR is APPROVED

This pull-request has been approved by: hakman, liranp

The full list of commands accepted by this bot can be found here.

The pull request process is described here

Needs approval from an approver in each of these files:

• pkg/model/spotinstmodel/OWNERS [hakman,liranp]

Approvers can indicate their approval by writing /approve in a comment
Approvers can cancel approval by writing /approve cancel in a comment

[ReillyTevera]

@liranp @hakman This should be cherry-picked into 1.19 as this issue also affects that version. This will result in Ocean instances not starting up correctly as they try to retrieve the instance group specification located at ig.ObjectMeta.Name + "." + b.ClusterName() instead of the actual spec located at ig.ObjectMeta.Name. Obviously, as that file does not exist the node is not able to bootstrap further.

[hakman]

The cherry-pick was trivial. If @liranp thinks it's safe, I don't mind.

[liranp]

@ReillyTevera, thank you for your feedback. @hakman, I believe we can do that.