Autoscaler: Generic API based cluster autoscaler
Despite being brought up in other issues #984 and #191 including with example implementations, I haven't seen this issue properly addressed. It would be awesome to have a cluster autoscaler that instead of interfacing directly with a cloud provider it would call a generic API endpoint with details about the request, for example to request a node to be removed or details about what kind of node to be created.
nikhilkhanna
All 8 comments
Thanks for opening this. Getting rid of cloud provider code from CA has clear advantages (no need to rebuild the image to update support for specific environment, for one.) Question is how to do this to 1) support most use-cases and 2) reduce, rather than increase, the burden of implementing and maintaining support for a new environment. This will probably be largely addressed by Cluster API project Instead of requiring the user to implement some CA specific API, we can then provide out-of-the-box support for it.
aleksandra-malinowska
on 10 Jul 2018
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
on 8 Oct 2018
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
fejta-bot
on 7 Nov 2018
Rotten issues close after 30d of inactivity.
Reopen the issue with /reopen.
Mark the issue as fresh with /remove-lifecycle rotten.
Send feedback to sig-testing, kubernetes/test-infra and/or fejta.
/close
fejta-bot
on 7 Dec 2018
@fejta-bot: Closing this issue.
In response to this:
Rotten issues close after 30d of inactivity.
Reopen the issue with/reopen.
Mark the issue as fresh with/remove-lifecycle rotten.Send feedback to sig-testing, kubernetes/test-infra and/or fejta.
/close
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.
k8s-ci-robot
on 7 Dec 2018
/reopen
/remove-lifecycle rotten
C'mon guys, monolith support of some evil cloud providers isn't fair.
Bessonov
on 11 Jan 2019
@Bessonov: You can't reopen an issue/PR unless you authored it or you are a collaborator.
In response to this:
/reopen
/remove-lifecycle rottenC'mon guys, monolith support of some evil cloud providers isn't fair.
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.
k8s-ci-robot
on 11 Jan 2019
There is an effort in progress to add support for Cluster API in cluster autoscaler, which would address this. There is still a lot of work to do - most importantly Cluster API lacks support for removing specific node (necessary for CA), but it's being worked on and should hopefully be available soon-ish.
MaciekPytel
on 14 Jan 2019
Related issues
dharmab
路
6Comments
johanneswuerbach
路
5Comments
clamoriniere
路
5Comments
tjliupeng
路
6Comments
adamrp
路
7Comments
Most helpful comment
There is an effort in progress to add support for Cluster API in cluster autoscaler, which would address this. There is still a lot of work to do - most importantly Cluster API lacks support for removing specific node (necessary for CA), but it's being worked on and should hopefully be available soon-ish.