deps: upgrade goceleris/loadgen to v1.0.0

cmd/bench/c2client.go

@@ -11,7 +11,8 @@ import (
 	"sync"
 	"time"
 
-	"github.com/goceleris/benchmarks/internal/bench"
+	"github.com/goceleris/loadgen"
+	"github.com/goceleris/loadgen/checkpoint"
 )
 
 // C2Client handles communication with the C2 orchestration server.
@@ -46,15 +47,15 @@ type C2BenchResult struct {
 	Errors         int64   `json:"errors"`
 
 	// System metrics — client-side and server-side resource usage.
-	ClientCPUPercent     float64                 `json:"client_cpu_percent,omitempty"`
-	ServerCPUPercent     float64                 `json:"server_cpu_percent,omitempty"`
-	ServerCPUUserPercent float64                 `json:"server_cpu_user_percent,omitempty"`
-	ServerCPUSysPercent  float64                 `json:"server_cpu_sys_percent,omitempty"`
-	ServerMemoryRSSMB    float64                 `json:"server_memory_rss_mb,omitempty"`
-	GCTotalPauseMs       float64                 `json:"gc_total_pause_ms,omitempty"`
-	GCMaxPauseMs         float64                 `json:"gc_max_pause_ms,omitempty"`
-	GCNumGC              int                     `json:"gc_num_gc,omitempty"`
-	Timeseries           []bench.TimeseriesPoint `json:"timeseries,omitempty"`
+	ClientCPUPercent     float64                   `json:"client_cpu_percent,omitempty"`
+	ServerCPUPercent     float64                   `json:"server_cpu_percent,omitempty"`
+	ServerCPUUserPercent float64                   `json:"server_cpu_user_percent,omitempty"`
+	ServerCPUSysPercent  float64                   `json:"server_cpu_sys_percent,omitempty"`
+	ServerMemoryRSSMB    float64                   `json:"server_memory_rss_mb,omitempty"`
+	GCTotalPauseMs       float64                   `json:"gc_total_pause_ms,omitempty"`
+	GCMaxPauseMs         float64                   `json:"gc_max_pause_ms,omitempty"`
+	GCNumGC              int                       `json:"gc_num_gc,omitempty"`
+	Timeseries           []loadgen.TimeseriesPoint `json:"timeseries,omitempty"`
 }
 
 // runHeartbeat sends periodic heartbeats to C2.
@@ -116,7 +117,7 @@ func (c *C2Client) sendHeartbeat(ctx context.Context) {
 }
 
 // sendResult sends a single benchmark result to C2.
-func (c *C2Client) sendResult(ctx context.Context, sr bench.ServerResult, result *bench.Result) error {
+func (c *C2Client) sendResult(ctx context.Context, sr checkpoint.ServerResult, result *loadgen.Result) error {
 	url := fmt.Sprintf("%s/api/worker/results", c.endpoint)
 
 	// Convert to C2 format with proper duration values

cmd/bench/config.go

@@ -10,7 +10,7 @@ import (
 	"strings"
 	"time"
 
-	"github.com/goceleris/benchmarks/internal/bench"
+	"github.com/goceleris/loadgen/checkpoint"
 )
 
 // benchConfig holds all parsed configuration for the benchmark runner.
@@ -333,22 +333,22 @@ func parseFlags() *benchConfig {
 }
 
 // setupCheckpoint creates or resumes a benchmark checkpoint.
-func setupCheckpoint(cfg *benchConfig) *bench.Checkpoint {
-	var checkpoint *bench.Checkpoint
+func setupCheckpoint(cfg *benchConfig) *checkpoint.Checkpoint {
+	var cp *checkpoint.Checkpoint
 	if cfg.resume {
-		if cp, err := bench.LoadCheckpoint(cfg.checkpointFile); err == nil {
-			checkpoint = cp
-			log.Printf("Resuming from checkpoint: %s (%d results completed)", cfg.checkpointFile, len(cp.Results))
+		if loaded, err := checkpoint.LoadCheckpoint(cfg.checkpointFile); err == nil {
+			cp = loaded
+			log.Printf("Resuming from checkpoint: %s (%d results completed)", cfg.checkpointFile, len(loaded.Results))
 		} else {
 			log.Printf("No existing checkpoint found, starting fresh")
 		}
 	}
 
-	if checkpoint == nil {
-		output := &bench.BenchmarkOutput{
+	if cp == nil {
+		output := &checkpoint.BenchmarkOutput{
 			Timestamp:    time.Now().UTC().Format("2006-01-02T15_04_05Z"),
 			Architecture: cfg.arch,
-			Config: bench.BenchmarkConfig{
+			Config: checkpoint.BenchmarkConfig{
 				Duration:      cfg.duration.String(),
 				DurationSecs:  cfg.duration.Seconds(),
 				WarmupSecs:    cfg.warmup.Seconds(),
@@ -359,9 +359,9 @@ func setupCheckpoint(cfg *benchConfig) *bench.Checkpoint {
 				H2MaxStreams:  cfg.h2Streams,
 				BodySizeBytes: defaultBodySize,
 			},
-			Results: []bench.ServerResult{},
+			Results: []checkpoint.ServerResult{},
 		}
-		checkpoint = bench.NewCheckpoint(output)
+		cp = checkpoint.NewCheckpoint(output)
 	}
 
 	if cfg.skipBenchmarks != "" {
@@ -374,7 +374,7 @@ func setupCheckpoint(cfg *benchConfig) *bench.Checkpoint {
 			}
 			parts := strings.SplitN(pair, ":", 2)
 			if len(parts) == 2 {
-				checkpoint.MarkCompleted(parts[0], parts[1])
+				cp.MarkCompleted(parts[0], parts[1])
 				skippedCount++
 			}
 		}
@@ -383,5 +383,5 @@ func setupCheckpoint(cfg *benchConfig) *bench.Checkpoint {
 		}
 	}
 
-	return checkpoint
+	return cp
 }

cmd/bench/runner.go

@@ -14,14 +14,16 @@ import (
 	"syscall"
 	"time"
 
-	"github.com/goceleris/benchmarks/internal/bench"
+	"github.com/goceleris/loadgen"
+	"github.com/goceleris/loadgen/checkpoint"
+
 	"github.com/goceleris/benchmarks/internal/dashboard"
 )
 
 // benchRunner coordinates benchmark execution across servers.
 type benchRunner struct {
 	cfg        *benchConfig
-	checkpoint *bench.Checkpoint
+	checkpoint *checkpoint.Checkpoint
 	rc         *RemoteController
 	c2         *C2Client
 }
@@ -150,19 +152,20 @@ func (r *benchRunner) runServerBenchmarks(ctx context.Context, serverType, serve
 		}
 
 		isH2 := strings.Contains(serverType, "-h2") || strings.Contains(serverType, "-hybrid")
-		cfg := bench.Config{
-			URL:           fmt.Sprintf("http://%s:%s%s", serverHost, r.cfg.port, bt.Path),
-			Method:        bt.Method,
-			Body:          bt.Body,
-			Headers:       bt.Headers,
-			Duration:      r.cfg.duration,
-			Connections:   r.cfg.connections,
-			Workers:       r.cfg.workers,
-			WarmupTime:    r.cfg.warmup,
-			KeepAlive:     true,
-			H2C:           isH2,
-			H2Connections: r.cfg.h2Conns,
-			H2MaxStreams:  r.cfg.h2Streams,
+		cfg := loadgen.Config{
+			URL:         fmt.Sprintf("http://%s:%s%s", serverHost, r.cfg.port, bt.Path),
+			Method:      bt.Method,
+			Body:        bt.Body,
+			Headers:     bt.Headers,
+			Duration:    r.cfg.duration,
+			Connections: r.cfg.connections,
+			Workers:     r.cfg.workers,
+			Warmup:      r.cfg.warmup,
+			HTTP2:       isH2,
+			HTTP2Options: loadgen.HTTP2Options{
+				Connections: r.cfg.h2Conns,
+				MaxStreams:  r.cfg.h2Streams,
+			},
 		}
 
 		// Fetch baseline server metrics before benchmark (remote mode only)
@@ -174,13 +177,17 @@ func (r *benchRunner) runServerBenchmarks(ctx context.Context, serverType, serve
 		}
 
 		// In remote mode, wrap the benchmark with retry logic
-		var result *bench.Result
+		var result *loadgen.Result
 		var err error
 
 		if r.cfg.remoteMode {
 			result, err = runBenchmarkWithRetry(ctx, cfg, r.rc, serverType, r.cfg.benchmarkTimeout, r.cfg.serverRetryInterval)
 		} else {
-			benchmarker := bench.New(cfg)
+			benchmarker, benchErr := loadgen.New(cfg)
+			if benchErr != nil {
+				log.Printf("ERROR: Failed to create benchmarker for %s/%s: %v — continuing with next benchmark", serverType, bt.Name, benchErr)
+				continue
+			}
 			result, err = benchmarker.Run(ctx)
 		}
 
@@ -204,21 +211,21 @@ func (r *benchRunner) runServerBenchmarks(ctx context.Context, serverType, serve
 
 		log.Printf("Completed: %.2f req/s, avg latency: %s", result.RequestsPerSec, result.Latency.Avg)
 
-		serverResult := result.ToServerResult(serverType, bt.Name, bt.Method, bt.Path)
+		serverResult := checkpoint.NewServerResult(result, serverType, bt.Name, bt.Method, bt.Path)
 
 		// Fetch post-benchmark server metrics and compute deltas
 		if r.cfg.remoteMode && r.rc != nil {
 			if postMetrics, err := r.rc.GetServerMetrics(ctx); err == nil {
 				sm := serverResult.Metrics
 				if sm == nil {
-					sm = &bench.SystemMetrics{}
+					sm = &checkpoint.SystemMetrics{}
 				}
 				sm.ServerCPUPercent = postMetrics.CPUPercent
 				sm.ServerCPUUserPercent = postMetrics.CPUUserPercent
 				sm.ServerCPUSysPercent = postMetrics.CPUSysPercent
 				sm.ServerMemoryRSSMB = postMetrics.MemoryRSSMB
 				if postMetrics.HasGC {
-					gc := &bench.GCPauseStats{
+					gc := &checkpoint.GCPauseStats{
 						TotalPauseMs: postMetrics.GCTotalPauseMs,
 						MaxPauseMs:   postMetrics.GCMaxPauseMs,
 						NumGC:        postMetrics.GCNumGC,
@@ -253,7 +260,7 @@ func (r *benchRunner) runServerBenchmarks(ctx context.Context, serverType, serve
 // saveResults writes final output files, dashboard JSON, and logs the summary.
 func (r *benchRunner) saveResults(completedBefore, skipped int) {
 	if r.cfg.mergeFile != "" {
-		if otherCP, err := bench.LoadCheckpoint(r.cfg.mergeFile); err == nil {
+		if otherCP, err := checkpoint.LoadCheckpoint(r.cfg.mergeFile); err == nil {
 			beforeMerge := len(r.checkpoint.Results)
 			r.checkpoint.MergeResults(otherCP)
 			afterMerge := len(r.checkpoint.Results)
@@ -326,7 +333,7 @@ func (r *benchRunner) saveResults(completedBefore, skipped int) {
 }
 
 // runBenchmarkWithRetry runs a benchmark with automatic retry on server failure.
-func runBenchmarkWithRetry(ctx context.Context, cfg bench.Config, rc *RemoteController, serverType string, timeout, interval time.Duration) (*bench.Result, error) {
+func runBenchmarkWithRetry(ctx context.Context, cfg loadgen.Config, rc *RemoteController, serverType string, timeout, interval time.Duration) (*loadgen.Result, error) {
 	deadline := time.Now().Add(timeout)
 	var lastErr error
 
@@ -383,7 +390,10 @@ func runBenchmarkWithRetry(ctx context.Context, cfg bench.Config, rc *RemoteCont
 		}
 
 		// Run the benchmark
-		benchmarker := bench.New(cfg)
+		benchmarker, benchErr := loadgen.New(cfg)
+		if benchErr != nil {
+			return nil, benchErr
+		}
 		result, err := benchmarker.Run(ctx)
 		if err != nil {
 			// Check if it's a connection error (server might have died)

go.mod

@@ -16,6 +16,7 @@ require (
 	github.com/gin-gonic/gin v1.12.0
 	github.com/go-chi/chi/v5 v5.2.5
 	github.com/goceleris/celeris v1.0.0
+	github.com/goceleris/loadgen v1.0.0
 	github.com/gofiber/fiber/v2 v2.52.12
 	github.com/google/uuid v1.6.0
 	github.com/hertz-contrib/http2 v0.1.8

go.sum

@@ -141,6 +141,8 @@ github.com/goccy/go-yaml v1.19.2 h1:PmFC1S6h8ljIz6gMRBopkjP1TVT7xuwrButHID66PoM=
 github.com/goccy/go-yaml v1.19.2/go.mod h1:XBurs7gK8ATbW4ZPGKgcbrY1Br56PdM69F7LkFRi1kA=
 github.com/goceleris/celeris v1.0.0 h1:MZsvsQVqck4lSxVom+6w02IHJva5hgwvydlMFtFTzP0=
 github.com/goceleris/celeris v1.0.0/go.mod h1:RiNVnUsmvfsPxGQaT7XQduPCF+amSn1AsipteFNO+2o=
+github.com/goceleris/loadgen v1.0.0 h1:3ViKRy+r1MUXtRvNrgWekZd1g9+9e+Yt/v43LnhUL2k=
+github.com/goceleris/loadgen v1.0.0/go.mod h1:/GSoaCJ2VR2Kt/nzfRLrJTXLrZN5lEF+bR9AqfzQRTc=
 github.com/gofiber/fiber/v2 v2.52.12 h1:0LdToKclcPOj8PktUdIKo9BUohjjwfnQl42Dhw8/WUw=
 github.com/gofiber/fiber/v2 v2.52.12/go.mod h1:YEcBbO/FB+5M1IZNBP9FO3J9281zgPAreiI1oqg8nDw=
 github.com/golang/protobuf v1.5.4 h1:i7eJL8qZTpSEXOPTxNKhASYpMn+8e5Q6AdndVa1dWek=

internal/dashboard/format.go

@@ -6,7 +6,8 @@ import (
 	"strings"
 	"time"
 
-	"github.com/goceleris/benchmarks/internal/bench"
+	"github.com/goceleris/loadgen/checkpoint"
+
 	"github.com/goceleris/benchmarks/internal/version"
 )
 
@@ -127,7 +128,7 @@ type Fastest struct {
 }
 
 // ConvertToDashboard converts a BenchmarkOutput to the dashboard JSON format.
-func ConvertToDashboard(output *bench.BenchmarkOutput, env EnvironmentInfo) *DashboardData {
+func ConvertToDashboard(output *checkpoint.BenchmarkOutput, env EnvironmentInfo) *DashboardData {
 	// Group results by benchmark type
 	suiteMap := make(map[string][]ServerBenchmark)
 	suiteOrder := []string{}
@@ -232,7 +233,7 @@ func ConvertToDashboard(output *bench.BenchmarkOutput, env EnvironmentInfo) *Das
 }
 
 // convertSystemMetrics converts bench.SystemMetrics to DashboardSystemMetrics.
-func convertSystemMetrics(m *bench.SystemMetrics) *DashboardSystemMetrics {
+func convertSystemMetrics(m *checkpoint.SystemMetrics) *DashboardSystemMetrics {
 	dm := &DashboardSystemMetrics{
 		ServerCPUPercent:     m.ServerCPUPercent,
 		ServerCPUUserPercent: m.ServerCPUUserPercent,
@@ -355,7 +356,7 @@ func parseCelerisObjective(server string) string {
 }
 
 // parseLatency converts string latency values to float64 milliseconds.
-func parseLatency(l bench.LatencyResult) DashboardLatency {
+func parseLatency(l checkpoint.LatencyResult) DashboardLatency {
 	return DashboardLatency{
 		Avg:   parseDurationMs(l.Avg),
 		Max:   parseDurationMs(l.Max),