Kudos, SonarCloud Quality Gate passed!   

0 Bugs
0 Vulnerabilities
0 Security Hotspots
0 Code Smells

No Coverage information
0.0% Duplication

Thanks for doing this! This is nitpicking, but it appears these benchmarks could be consolidated as sub-benchmarks parameterized by user and token counts i.e., something like this:

func BenchmarkCache(b *testing.B) {
	for _, bm := range []struct {
		users, tokens int
	}{
		{100, 10_000},
	} {
		b.Run(fmt.Sprintf("%d users %d tokens", bm.users, bm.tokens), func(b *testing.B) {
			// inline benchmark() and perhaps also queryCache()
		})
	}
}

...which would be tidier. I haven't tried it though, maybe there's a catch. What do you think?

Great idea, I agree that’s better

Would be nice to integrate with https://github.com/benchmark-action/github-action-benchmark

Do we still want to merge this in?

I wouldn't block anything on it but I think we should merge this (after tidying it up into sub-benchmarks as I described above). It's weird to benchmark with a unit test printing to console instead of go's standard benchmarking framework.

Kudos, SonarCloud Quality Gate passed!   

0 Bugs
0 Vulnerabilities
0 Security Hotspots
0 Code Smells

No Coverage information
0.3% Duplication

after tidying it up into sub-benchmarks

@chlowell Done, ready for review

Kudos, SonarCloud Quality Gate passed!   

0 Bugs
0 Vulnerabilities
0 Security Hotspots
0 Code Smells

No Coverage information
0.3% Duplication