uses monoprocess backend for bench

The new polling interval introduces much deviation
on the tests, plus it makes the test unnecessarily
slow. This should help visualizing better
the performance of each individual backend.

Author: vroldanbet

bench/main.go

@@ -12,6 +12,7 @@ import (
 	"time"
 
 	"github.com/cschleiden/go-workflows/backend"
+	"github.com/cschleiden/go-workflows/backend/monoprocess"
 	"github.com/cschleiden/go-workflows/backend/mysql"
 	"github.com/cschleiden/go-workflows/backend/redis"
 	"github.com/cschleiden/go-workflows/backend/sqlite"
@@ -103,12 +104,12 @@ func main() {
 func getBackend(b string, opt ...backend.BackendOption) backend.Backend {
 	switch b {
 	case "memory":
-		return sqlite.NewInMemoryBackend(opt...)
+		return monoprocess.NewMonoprocessBackend(sqlite.NewInMemoryBackend(opt...))
 
 	case "sqlite":
 		os.Remove("bench.sqlite")
 
-		return sqlite.NewSqliteBackend("bench.sqlite", opt...)
+		return monoprocess.NewMonoprocessBackend(sqlite.NewSqliteBackend("bench.sqlite", opt...))
 
 	case "mysql":
 		db, err := sql.Open("mysql", fmt.Sprintf("%s:%s@/?parseTime=true&interpolateParams=true", "root", "root"))
@@ -128,7 +129,7 @@ func getBackend(b string, opt ...backend.BackendOption) backend.Backend {
 			panic(err)
 		}
 
-		return mysql.NewMysqlBackend("localhost", 3306, "root", "root", "bench", opt...)
+		return monoprocess.NewMonoprocessBackend(mysql.NewMysqlBackend("localhost", 3306, "root", "root", "bench", opt...))
 
 	case "redis":
 		rclient := redisv8.NewUniversalClient(&redisv8.UniversalOptions{