Autoscaler: Provide PricingModel for AWS

Created on 20 Jun 2017 · 14Comments · Source: kubernetes/autoscaler

Follow up of #82

cc: @mumoshu

areprovideaws cluster-autoscaler help wanted lifecyclrotten

Source

mwielgus

👍9

Most helpful comment

@mrcrgl Absolutely, please send us a PR.

mwielgus on 7 Nov 2017

❤4

All 14 comments

We're interested in this feature. Are contributions welcome?

mrcrgl on 7 Nov 2017

@mrcrgl Absolutely, please send us a PR.

mwielgus on 7 Nov 2017

❤4

@mwielgus Ok. I'll start on it. Is it fine for you guys to evaluate the spot prices on demand via the spot price history api? I also tend to load the price list on demand as well (with caching).

mrcrgl on 16 Nov 2017

@mrcrgl It looks like some of the work for the on-demand pricing was started here:

https://github.com/kubernetes/autoscaler/blob/master/cluster-autoscaler/cloudprovider/aws/ec2_instance_types/gen.go

Because pricing can be dependent on your region and account features, this seems to cover some of that, but it needs to be fixed to work with the current API response formats.

Support for detecting the GPU field and instance types would also be nice for that.

I think getting the spot prices from the spot price history API makes perfect sense.

kylegato on 21 Nov 2017

👍1

I've worked on it. I'll test these feature within the next few days on one of our clusters.
Does this change work for you guys?

mrcrgl on 22 Nov 2017

@mwielgus PTAL https://github.com/kubernetes/autoscaler/pull/486

mrcrgl on 1 Dec 2017

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 1 Mar 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
/remove-lifecycle stale

fejta-bot on 31 Mar 2018

The lifecycle/rotten label should be removed; the PR is still active.

patrickf55places on 27 Apr 2018

/remove-lifecycle stale

aleksandra-malinowska on 21 May 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 20 Jun 2018

/reopen

bskiba on 21 Jun 2018

@bskiba: you can't re-open an issue/PR unless you authored it or you are assigned to it.

In response to this:

/reopen

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.