[DISCO-3010] Calibrate smoke and average tests for n2-highcpu-8 machine change and real-world network traffic data [load test: warn]

References

JIRA: DISCO-3010

Description

update network traffic end point distribution from 50% suggest and 50% recommendations to 65% suggest and 35% recommendations
add corrections and clarifications to documentation

For the latest load test results, please see the Merino Load Test Spreadsheet

Calibrating for n2-highcpu-8 machine change

In the end, no additional load was necessary, since scaling is still achieved with the configured load

Adjusting traffic distribution

The current load distribution between endpoints has proven to be slightly different than the original 50/50 estimate. So adjustments have been made as follows:

49.7% --> 64.2% weatherSuggestions 42.1% --> 29.8% curatedRecommendationsLocale 7.4% --> 5.2% curatedRecommendationsRandomTopics 0.2% --> 0.2% admSuggestions 0.2% --> 0.2% dynamicWikipediaSuggestions 0.2% --> 0.2% fakerSuggestions 0.2% --> 0.2% topPicksSuggestions

PR Review Checklist

Put an x in the boxes that apply

[x] This PR conforms to the Contribution Guidelines
[x] The PR title starts with the JIRA issue reference, format [DISCO-####], and has the same title (if applicable)
[x] [load test: (abort|warn)] keywords are applied (if applicable)
[x] Documentation has been updated (if applicable)
[x] Functional and performance test coverage has been expanded and maintained (if applicable)

mozilla-services / merino-py