DeterministicSamplerTest flakey

[vreynolds]

I noticed that the DeterministicSamplerTest fails occasionally on dependabot PRs (for example: https://app.circleci.com/pipelines/github/honeycombio/beeline-go/958/workflows/37be948a-6bb5-4606-a629-e5c6f785e07e/jobs/2992)

Re-running passes

[vreynolds]

Looks like the failure is always just a little off the margin of error (~10 or less off), and always with a sample rate of 100. This makes sense, because with 200 000 test cases and sample rate of 100, the test expects to see results in the 2000 (200 000/100) +/- 100 (2000 * 0.05) range. For comparison, with a sample rate of 50, the test expects to see results in the 4000 (200 000/50) +/- 200 (4000 * 0.05) range. (Smaller sample rate => larger window of error)

I think the options are:

• Increase the margin of error for the 100 sample rate
• Increase the total number of test cases (200k currently)
• Don't test the 100 sample rate (I'm leaning towards this, because I noticed that the Go beeline is an outlier, and other ones don't test sample rates higher than 50)

I see we already ran into this in the Ruby beeline in the past: honeycombio/beeline-ruby#107 and decided to increase the margin of error for a higher sample rate.