Skip to content

README.md

Latest commit

 

History

History
107 lines (71 loc) · 3.24 KB

README.md

File metadata and controls

107 lines (71 loc) · 3.24 KB

blocking-testing

Module purpose

blocking-testing contains Cucumber end-to-end tests for the Blocking Echo system.
It validates behavior across real client/server classes from:

  • blocking-client
  • blocking-server

This module is test-focused and does not produce runtime application code.

What is covered

Feature files:

  • src/test/resources/features/blocking.feature (functional + edge cases + high load)
  • src/test/resources/features/blocking-benchmark.feature (performance-only scenarios)

Main coverage areas:

  • Simple connect/send/disconnect flows.
  • Multi-client counter consistency (stats).
  • Message format and edge cases (empty input, spaces, escaped content).
  • High-load behavior (100 clients x 100 messages each).
  • Performance benchmark scenarios for server/client thread combinations.

Step definition classes:

  • BlockingStepDefinitions - functional and load steps.
  • BlockingBenchmarkStepDefinitions - benchmark/performance steps.
  • CommonStepDefinitions - shared steps and cleanup hook.

Tag model

Functional scenarios:

  • @Virtual - functional flows using virtual clients.
  • @Platform - functional flows using platform clients.

Performance scenarios:

  • @Performance - benchmark scenarios.
  • @ServerPlatform / @ServerVirtual - server type.
  • @ClientPlatform / @ClientVirtual - client type.
  • @PerformanceSummary - compares persisted benchmark outputs from all 4 combinations.

Benchmark design

Implemented in BlockingBenchmarkStepDefinitions (with shared client operations from BlockingStepDefinitions):

  • warmup iterations (not included in measured average/median)
  • measured runs
  • per-run metrics:
    • elapsed time
    • throughput (msg/s)
    • average latency (ms/msg)
  • persisted result files under target/performance-results/*.properties
  • final summary step reads all persisted files, sorts by average time, and prints:
    • fastest to slowest
    • absolute delta and percent delta
    • total measured time and total scenario time

During benchmark message loops, step definitions temporarily suppress System.out noise for more stable timings.

How tests are run

Runner: dev.nklip.javacraft.echo.blocking.CucumberRunner

Reports generated to:

  • target/cucumber-reports/cucumber.html
  • target/cucumber-reports/cucumber.json

Execute from project root

Run full cucumber suite for this module:

mvn -pl echo/blocking/blocking-testing -am test

Run only functional virtual scenarios:

mvn -pl echo/blocking/blocking-testing -am test -Dcucumber.filter.tags='@Virtual and not @Performance and not @PerformanceSummary'

Run only functional platform scenarios:

mvn -pl echo/blocking/blocking-testing -am test -Dcucumber.filter.tags='@Platform and not @Performance and not @PerformanceSummary'

Run one benchmark combination example (PlatformServer + PlatformThreadClient):

mvn -pl echo/blocking/blocking-testing -am test -Dcucumber.filter.tags='@Performance and @ServerPlatform and @ClientPlatform'

After running all 4 @Performance combinations, run final comparison:

mvn -pl echo/blocking/blocking-testing -am test -Dcucumber.filter.tags='@PerformanceSummary'

If @PerformanceSummary is run before benchmark result files exist, the test fails by design.