Run provider getters in parallel to reduce startup time

[michaelasp]

When reviewing #116 I saw that we don't run these getters in parallel. Since all of them time out after ~5 seconds, worst case scenario we take 5*(# of providers) to start up. Run these in parallel, the network overhead should be minimal and we already reduce the worst case startup time by 10 seconds.

[k8s-ci-robot]

[APPROVALNOTIFIER] This PR is APPROVED

This pull-request has been approved by: michaelasp

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

The pull request process is described here

Details

Needs approval from an approver in each of these files:

• OWNERS [michaelasp]

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

[aojea]

can we identify if the urls are the same? running a bunch of http requests in parallel will hit some rate limit and make it worse ... also we should have a metrics for startup latency and one that specifically measure this time to identify the instance ...

I want to avoid premature optimization here

[michaelasp]

can we identify if the urls are the same? running a bunch of http requests in parallel will hit some rate limit and make it worse ... also we should have a metrics for startup latency and one that specifically measure this time to identify the instance ...

I want to avoid premature optimization here

169.254.169.254 is the backend for all of them, so likely we will get a response, but we will poll until timeout on them all from what I can tell.

Let me take a look, no matter what just looking at it heuristically, if we use OKE today then we have to time out twice on startup. 10s isn't too bad but will increase when we add AWS, Alibaba, etc...

I can add startup latency metrics just to see whether this hypothesis is correct.