Test scenarios for validating box plots

[chandrams]

Test scenarios for validating box plots

• Update functional tests to validate plot data

  • [x] Validate data points count
  • [x] Validate if the actual count of data displayed matches the expected data points count
  • [x] Test with minimal data points for each term and validate the plots section
  • [x] Ensure plots are not displayed when there are no min usage metrics present for each of the terms (for example, short term min metrics is 2)
  • [x] Test box plots with the below combination of recommendation terms with contiguous and non-contiguous usage data
  • ('L', 'M', 'S')
  • ('L', 'M', 'NS')
  • ('L', 'NM', 'S')
  • ('L', 'NM', 'NS')
  • ('NL', 'M', 'S')
  • ('NL', 'M', 'NS')
  • ('NL', 'NM', 'S')
  • ('NL', 'NM', 'NS') Where NL - No long term, NM - No medium term, NS - No short term
  • [x] Validate the box plot values (min, q1, median, q3, max, format)
  • Is there a flag to switch off plot data?

• Scalability runs to capture resource usage

  • [x] Scalability run - 5k / 15 days (Check for DB Resource usage before & after using box plots)
  • [ ] Scalability run - 100k /15 days (Check for DB Resource usage before & after using box plots)
  • [ ] Test With older data + scalability run

• Kruize UI validation

  • [ ] Validate box plots on Kruize UI:
  • [ ] with all data points present for all terms
  • [ ] With missing data points for all terms
  • [ ] [6-12] hour data missing
  • [ ] Data missing for any of the days for medium or long term
  • [ ] Test with a few exps having older recommendations (no box plot data) and a few having box plot data

• Kruize demos

  • [ ] Update the Validation of actual recommendations to validate box plot values

• [ ] Documentation validation

• [ ] Validate documentation on box plots

[rbadagandi1]

Closing this issue since this got merged as part of 0.0.22