Add PartitionInstanceRing.GetReplicationSetForPartitionAndOperation()

[pracucci]

What this PR does:

Add PartitionInstanceRing.GetReplicationSetForPartitionAndOperation(). This PR is against usage-tracker branch, which is a branch me and @colega are using to build a prototype. This will not be merged to main for now.

Notes to reviewers:

• Review with "hide whitespace changes"

Ensuring no regression on the already-existing GetReplicationSetsForOperation():

goos: darwin
goarch: arm64
pkg: github.com/grafana/dskit/ring
cpu: Apple M3 Pro
                                                        │ before.txt  │             after.txt              │
                                                        │   sec/op    │   sec/op     vs base               │
PartitionInstanceRing_GetReplicationSetsForOperation-11   21.81µ ± 0%   22.42µ ± 1%  +2.80% (p=0.000 n=10)

                                                        │  before.txt  │            after.txt             │
                                                        │     B/op     │     B/op      vs base            │
PartitionInstanceRing_GetReplicationSetsForOperation-11   39.12Ki ± 0%   39.12Ki ± 0%  ~ (p=1.000 n=10) ¹
¹ all samples are equal

                                                        │ before.txt │           after.txt            │
                                                        │ allocs/op  │ allocs/op   vs base            │
PartitionInstanceRing_GetReplicationSetsForOperation-11   101.0 ± 0%   101.0 ± 0%  ~ (p=1.000 n=10) ¹
¹ all samples are equal

Which issue(s) this PR fixes:

N/A

Checklist

• Tests updated
• CHANGELOG.md updated - the order of entries should be [CHANGE], [FEATURE], [ENHANCEMENT], [BUGFIX]

[colega]

I think we can avoid the allocation here completely.

Suggested change

	zonesBuffer := make([]string, 0, 3) // Pre-allocate buffer assuming 3 zones.
	var stackZonesBuffer [3]string // Pre-allocate buffer assuming 3 zones.
	zonesBuffer := stackZonesBuffer[:]

(I'd also do 5, not 3, let's not be greedy)

[pracucci]

Picked the on-stack buffer: bf3dc4a. About 3 vs 5 I will leave it for a separate discussion, given in the ring/ pkg we assume 3 in different places.

[colega]

LGTM, much easier to review with whitespace ignored.