Gruptor is a basic port of Disruptor (JAVA) . It's very fast by using lock-free ringbuffer
const DefaultBufferSize = 1024 // must be 2,4,8,..1024...
const DefaultBufferMask = DefaultBufferSize -1
var ringBuffer [DefaultBufferSize]int64 // specified an ringBuffer
// implements your own Consumer
type AConsumer struct {
}
func (c *AConsumer) Consume(lo, hi int64) {
for lo <= hi {
event := ringBuffer[lo&DefaultBufferMask]
if event != lo {
warning := fmt.Sprintf("\nRace condition--Sequence: %d, Event: %d\n", lo, event)
fmt.Printf(warning)
panic(warning)
}
lo++
}
}
// Create the gruptor
g := gruptor.NewBuilder(DefaultBufferSize).HandleEventWith(&AConsumer{}).Build()
g.Start()
defer g.Stop()
// using Writer to acquiring next sequence
w := g.Writer()
var sequence int64
for sequence < 100000 {
sequence = w.Next()
ringBuffer[sequence&DefaultBufferMask] = sequence
w.Commit(sequence, sequence)
}// TODO add some examplegoos: darwin
goarch: amd64
pkg: gruptor/benchmarks
BenchmarkCompositeBarrierWithOneCursor_Read-8 2000000000 1.82 ns/op 0 B/op 0 allocs/op
BenchmarkCompositeBarrierWithMoreCursor_Read-8 200000000 6.35 ns/op 0 B/op 0 allocs/op
BenchmarkBlockingOneGoroutine-8 30000000 55.1 ns/op 0 B/op 0 allocs/op
BenchmarkBlockingTwoGoroutines-8 20000000 84.1 ns/op 0 B/op 0 allocs/op
BenchmarkBlockingThreeGoroutinesWithContendedWrite-8 10000000 167 ns/op 0 B/op 0 allocs/op
BenchmarkUnBlockingOneGoroutine-8 100000000 16.9 ns/op 0 B/op 0 allocs/op
BenchmarkUnBlockingTwoGoroutines-8 100000000 17.4 ns/op 0 B/op 0 allocs/op
BenchmarkUnBlockingThreeGoroutinesWithContendedWrite-8 100000000 63.1 ns/op 0 B/op 0 allocs/op
BenchmarkCursor_Load-8 2000000000 0.81 ns/op 0 B/op 0 allocs/op
BenchmarkCursor_Store-8 2000000000 0.80 ns/op 0 B/op 0 allocs/op
BenchmarkCursor_Read-8 2000000000 0.83 ns/op 0 B/op 0 allocs/op
BenchmarkRingBuffer_Get-8 500000000 2.92 ns/op 0 B/op 0 allocs/op
BenchmarkRingBuffer_Published-8 500000000 2.90 ns/op 0 B/op 0 allocs/op
BenchmarkRingBuffer_Set-8 30000000 52.0 ns/op 8 B/op 1 allocs/op
BenchmarkCustomRingBuffer_Get-8 1000000000 2.25 ns/op 0 B/op 0 allocs/op
BenchmarkCustomRingBuffer_Set-8 500000000 3.00 ns/op 0 B/op 0 allocs/op
BenchmarkCustomRingBufferWithEvent_Get-8 500000000 3.89 ns/op 0 B/op 0 allocs/op
BenchmarkCustomRingBufferWithEvent_Set-8 30000000 55.7 ns/op 8 B/op 1 allocs/op
BenchmarkNoTypeSwitch-8 2000000000 1.04 ns/op 0 B/op 0 allocs/op
BenchmarkUsingTypeSwitch-8 2000000000 1.52 ns/op 0 B/op 0 allocs/op
BenchmarkUsingTypeAssertion-8 1000000000 2.29 ns/op 0 B/op 0 allocs/op
BenchmarkGruptor_OneWriterOneConsumer-8 100000000 19.7 ns/op 0 B/op 0 allocs/op
BenchmarkGruptor_OneWriterOneConsumerMoreCPU-8 50000000 32.3 ns/op 0 B/op 0 allocs/op
BenchmarkGruptor_OneWriterMultiConsumer-8 50000000 32.8 ns/op 0 B/op 0 allocs/op
BenchmarkGruptor_MultiWriterOneConsumer-8 10000000 168 ns/op 0 B/op 0 allocs/op
BenchmarkGruptor_MultiWriterMultiConsumer-8 10000000 184 ns/op 0 B/op 0 allocs/op
BenchmarkGruptor_MultiWriterOneConsumerInMultiGoroutines-8 5000000 241 ns/op 0 B/op 0 allocs/op