Send non-reachability DescribeTaskQueue requests through frontend rate limiter

[carlydf]

What changed?

Previously, all DescribeTaskQueue requests use visibility rate limiter, which is too restrictive for non-reachability requests. Fix that by checking in the interceptor.

Why?

How did you test it?

Potential risks

Documentation

Is hotfix candidate?

[ShahabT]

I think we need to follow up this quickly by adding caching for the stats otherwise I'm concerned about the fanout load that each describetaskqueue request will make.

[carlydf]

I think we need to follow up this quickly by adding caching for the stats otherwise I'm concerned about the fanout load that each describetaskqueue request will make.

Should @Shivs11 work on that? So far only one person has complained about the DescribeTaskQueue rate limit, so we could hold off on merging this PR until we are confident that task queue stats are protected. Mostly I wanted to make sure that the reachability rate limit would not impact users' experience of getting ApproximateBacklogCount

[Shivs11]

I think we need to follow up this quickly by adding caching for the stats otherwise I'm concerned about the fanout load that each describetaskqueue request will make.

Should @Shivs11 work on that? So far only one person has complained about the DescribeTaskQueue rate limit, so we could hold off on merging this PR until we are confident that task queue stats are protected. Mostly I wanted to make sure that the reachability rate limit would not impact users' experience of getting ApproximateBacklogCount

Agreed, lets hold off on getting this PR in. I shall work on getting caching added for the stats