disable default monitoring

[ElvinEfendi]

What this PR does / why we need it:
We don't use any stats from default monitoring. And we know that it's causing the controller to use a lot of resource: kubernetes#3314

This PR disables that.

Which issue this PR fixes (optional, in fixes #<issue number>(, fixes #<issue_number>, ...) format, will close that issue when PR gets merged): fixes #

Special notes for your reviewer:

@Shopify/edgescale

[ElvinEfendi]

Image released index.docker.io/shopify/nginx-ingress-controller:0.21.0-no-monitor

[ElvinEfendi]

I deployed this to tierstaging-central just to make sure it does not break anything and it seems to be working fine

[wayt]

Can we close this now? Or should we make this an option on our ingress ?

[ElvinEfendi]

@wayt I'm leaning toward merging this. Any objection? The current plan for upstream is to introduce a flag to disable this monitoring, but until that happens we can merge this PR IMO.

[ElvinEfendi]

Another reason to merge this PR: https://logs.shopify.io/en-US/app/search/report?sid=_cGVlci5hbGxhbg_cGVlci5hbGxhbg__search__RMD5351807261e537e5c_at_1543437786_1468_6376F8BF-5E5C-4E4C-9BF8-1E079BB8768F&s=%2FservicesNS%2Fnobody%2Fsearch%2Fsaved%2Fsearches%2FDataDog%2520Top%2520Metrics%2520with%2520Cost

If you ctrl+f for ingress_nginx in that page you'll see we spends thousands $ per month.

That being said I'm surprised to see them there because we have disabled DD Prometheus checks.

[wayt]

@ElvinEfendi 👍 we're using statsd instead of this right?

[ElvinEfendi]

we're using statsd instead of this right?

correct