Since benchmarks are meant to test in a consistent environment, we cannot execute the benchmark on RBE executors as executors do not run in calibrated environments. PR Close #34996
		
			
				
	
	
		
			30 lines
		
	
	
		
			913 B
		
	
	
	
		
			Python
		
	
	
	
	
	
			
		
		
	
	
			30 lines
		
	
	
		
			913 B
		
	
	
	
		
			Python
		
	
	
	
	
	
load("//tools:defaults.bzl", "protractor_web_test_suite")
 | 
						|
 | 
						|
"""
 | 
						|
  Macro that can be used to define a benchmark test. This differentiates from
 | 
						|
  a normal Protractor test suite because we specify a custom "perf" configuration
 | 
						|
  that sets up "@angular/benchpress". Benchmark test targets will not run on CI
 | 
						|
  unless explicitly requested.
 | 
						|
"""
 | 
						|
 | 
						|
def benchmark_test(name, server, tags = [], **kwargs):
 | 
						|
    protractor_web_test_suite(
 | 
						|
        name = name,
 | 
						|
        configuration = "//:protractor-perf.conf.js",
 | 
						|
        data = [
 | 
						|
            "//packages/benchpress",
 | 
						|
        ],
 | 
						|
        on_prepare = "//modules/benchmarks:start-server.js",
 | 
						|
        server = server,
 | 
						|
        # Benchmark targets should not run on CI by default.
 | 
						|
        tags = tags + [
 | 
						|
            "manual",
 | 
						|
            "no-remote-exec",
 | 
						|
        ],
 | 
						|
        test_suite_tags = [
 | 
						|
            "manual",
 | 
						|
            "no-remote-exec",
 | 
						|
        ],
 | 
						|
        **kwargs
 | 
						|
    )
 |