benchmarks

Documentation

Index

• Constants
• Variables
• func NewArwenBenchmark(arg ArgArwenBenchmark) *arwenBenchmark
• func NewCoordinator(benchmarks []BenchmarkRunner) (*coordinator, error)
• func NewDelegationBenchmark(arg ArgDelegationBenchmark) *delegationBenchmark
• func NewErc20Benchmark(arg ArgErc20Benchmark) *erc20Benchmark
• type ArgArwenBenchmark
• type ArgDelegationBenchmark
• type ArgErc20Benchmark
• type BenchmarkRunner
• type SingleResult
• type TestResults
  • func (tr *TestResults) ToDisplayTable() string
  • func (tr *TestResults) ToStrings() [][]string

Constants

const ThresholdEnoughComputingPower = 31 * time.Second

ThresholdEnoughComputingPower represents the threshold considered when deciding if a Node Under Test (NUT) is considered having (or not) enough compute capability when executing SC calls. if the result.TotalDuration exceeds this const, the NUT is considered less powerful (the NUT needed more time to process same amount of work). The value is hardcoded and was decided through testing.

Variables

var ErrEmptyBenchmarksSlice = errors.New("empty benchmarks slice provided")

ErrEmptyBenchmarksSlice signals that the provided benchmarks slice was empty

var ErrFileDoesNotExist = errors.New("file does not exist")

ErrFileDoesNotExist signals that the required file does not exist

var ErrNilBenchmark = errors.New("nil benchmark")

ErrNilBenchmark signals that a nil benchmark was provided

Functions

func NewArwenBenchmark

func NewArwenBenchmark(arg ArgArwenBenchmark) *arwenBenchmark

NewArwenBenchmark creates a new benchmark based on SC execution through Arwen VM

func NewCoordinator

func NewCoordinator(benchmarks []BenchmarkRunner) (*coordinator, error)

NewCoordinator will create a coordinator used to launch all provided benchmarks

func NewDelegationBenchmark

func NewDelegationBenchmark(arg ArgDelegationBenchmark) *delegationBenchmark

NewDelegationBenchmark creates a new benchmark based on delegation SC execution through Arwen VM

func NewErc20Benchmark

func NewErc20Benchmark(arg ArgErc20Benchmark) *erc20Benchmark

NewErc20Benchmark creates a new benchmark based on erc20 SC execution through Arwen VM

Types

type ArgArwenBenchmark

type ArgArwenBenchmark struct {
	Name         string
	ScFilename   string
	TestingValue uint64
	Function     string
	Arguments    [][]byte
	NumRuns      int
}

ArgArwenBenchmark is the Arwen type benchmark argument used in constructor

type ArgDelegationBenchmark

type ArgDelegationBenchmark struct {
	Name               string
	ScFilename         string
	NumRuns            int
	NumBatches         uint32
	NumTxPerBatch      uint32
	NumQueriesPerBatch uint32
}

ArgDelegationBenchmark is the delegation type benchmark argument used in constructor

type ArgErc20Benchmark

type ArgErc20Benchmark struct {
	Name               string
	ScFilename         string
	Function           string
	NumRuns            int
	NumTransfersPerRun int
}

ArgErc20Benchmark is the erc20 type benchmark argument used in constructor

type BenchmarkRunner

type BenchmarkRunner interface {
	Run() (time.Duration, error)
	Name() string
	IsInterfaceNil() bool
}

BenchmarkRunner defines a benchmark test able to measure in some way the host

type SingleResult

type SingleResult struct {
	time.Duration
	Name  string
	Error error
}

SingleResult contains the output data after a benchmark run

type TestResults

type TestResults struct {
	TotalDuration        time.Duration
	Error                error
	Results              []SingleResult
	EnoughComputingPower bool
}

TestResults represents the output structure containing the test results data

func (*TestResults) ToDisplayTable

func (tr *TestResults) ToDisplayTable() string

ToDisplayTable will output the contained data as an ASCII table

func (*TestResults) ToStrings

func (tr *TestResults) ToStrings() [][]string

ToStrings will return the contained data as strings (to be easily written, e.g. in a file)