remove memory estimation logic and related structures

Author: ilopezluna

main.go

@@ -11,13 +11,11 @@ import (
 	"syscall"
 	"time"
 
-	"github.com/docker/model-runner/pkg/gpuinfo"
 	"github.com/docker/model-runner/pkg/inference"
 	"github.com/docker/model-runner/pkg/inference/backends/llamacpp"
 	"github.com/docker/model-runner/pkg/inference/backends/mlx"
 	"github.com/docker/model-runner/pkg/inference/backends/vllm"
 	"github.com/docker/model-runner/pkg/inference/config"
-	"github.com/docker/model-runner/pkg/inference/memory"
 	"github.com/docker/model-runner/pkg/inference/models"
 	"github.com/docker/model-runner/pkg/inference/scheduling"
 	"github.com/docker/model-runner/pkg/metrics"
@@ -65,15 +63,6 @@ func main() {
 		llamaServerPath = "/Applications/Docker.app/Contents/Resources/model-runner/bin"
 	}
 
-	gpuInfo := gpuinfo.New(llamaServerPath)
-
-	sysMemInfo, err := memory.NewSystemMemoryInfo(log, gpuInfo)
-	if err != nil {
-		log.Fatalf("unable to initialize system memory info: %v", err)
-	}
-
-	memEstimator := memory.NewEstimator(sysMemInfo)
-
 	// Create a proxy-aware HTTP transport
 	// Use a safe type assertion with fallback, and explicitly set Proxy to http.ProxyFromEnvironment
 	var baseTransport *http.Transport
@@ -93,7 +82,6 @@ func main() {
 		log,
 		clientConfig,
 		nil,
-		memEstimator,
 	)
 	modelManager := models.NewManager(log.WithFields(logrus.Fields{"component": "model-manager"}), clientConfig)
 	log.Infof("LLAMA_SERVER_PATH: %s", llamaServerPath)
@@ -118,12 +106,6 @@ func main() {
 		log.Fatalf("unable to initialize %s backend: %v", llamacpp.Name, err)
 	}
 
-	if os.Getenv("MODEL_RUNNER_RUNTIME_MEMORY_CHECK") == "1" {
-		memory.SetRuntimeMemoryCheck(true)
-	}
-
-	memEstimator.SetDefaultBackend(llamaCppBackend)
-
 	vllmBackend, err := vllm.New(
 		log,
 		modelManager,
@@ -160,7 +142,6 @@ func main() {
 			"",
 			false,
 		),
-		sysMemInfo,
 	)
 
 	// Create the HTTP handler for the scheduler

pkg/inference/memory/estimator.go

@@ -17,23 +17,10 @@ import (
 	"github.com/docker/model-runner/pkg/distribution/builder"
 	reg "github.com/docker/model-runner/pkg/distribution/registry"
 	"github.com/docker/model-runner/pkg/inference"
-	"github.com/docker/model-runner/pkg/inference/memory"
 
 	"github.com/sirupsen/logrus"
 )
 
-type mockMemoryEstimator struct{}
-
-func (me *mockMemoryEstimator) SetDefaultBackend(_ memory.MemoryEstimatorBackend) {}
-
-func (me *mockMemoryEstimator) GetRequiredMemoryForModel(_ context.Context, _ string, _ *inference.BackendConfiguration) (inference.RequiredMemory, error) {
-	return inference.RequiredMemory{RAM: 0, VRAM: 0}, nil
-}
-
-func (me *mockMemoryEstimator) HaveSufficientMemoryForModel(_ context.Context, _ string, _ *inference.BackendConfiguration) (bool, inference.RequiredMemory, inference.RequiredMemory, error) {
-	return true, inference.RequiredMemory{}, inference.RequiredMemory{}, nil
-}
-
 // getProjectRoot returns the absolute path to the project root directory
 func getProjectRoot(t *testing.T) string {
 	// Start from the current test file's directory
@@ -123,11 +110,10 @@ func TestPullModel(t *testing.T) {
 	for _, tt := range tests {
 		t.Run(tt.name, func(t *testing.T) {
 			log := logrus.NewEntry(logrus.StandardLogger())
-			memEstimator := &mockMemoryEstimator{}
 			handler := NewHTTPHandler(log, ClientConfig{
 				StoreRootPath: tempDir,
 				Logger:        log.WithFields(logrus.Fields{"component": "model-manager"}),
-			}, nil, memEstimator)
+			}, nil)
 
 			r := httptest.NewRequest(http.MethodPost, "/models/create", strings.NewReader(`{"from": "`+tag+`"}`))
 			if tt.acceptHeader != "" {
@@ -234,13 +220,12 @@ func TestHandleGetModel(t *testing.T) {
 	for _, tt := range tests {
 		t.Run(tt.name, func(t *testing.T) {
 			log := logrus.NewEntry(logrus.StandardLogger())
-			memEstimator := &mockMemoryEstimator{}
 			handler := NewHTTPHandler(log, ClientConfig{
 				StoreRootPath: tempDir,
 				Logger:        log.WithFields(logrus.Fields{"component": "model-manager"}),
 				Transport:     http.DefaultTransport,
 				UserAgent:     "test-agent",
-			}, nil, memEstimator)
+			}, nil)
 
 			// First pull the model if we're testing local access
 			if !tt.remote && !strings.Contains(tt.modelName, "nonexistent") {
@@ -315,11 +300,10 @@ func TestCors(t *testing.T) {
 	for _, tt := range tests {
 		t.Run(tt.path, func(t *testing.T) {
 			t.Parallel()
-			memEstimator := &mockMemoryEstimator{}
 			discard := logrus.New()
 			discard.SetOutput(io.Discard)
 			log := logrus.NewEntry(discard)
-			m := NewHTTPHandler(log, ClientConfig{}, []string{"*"}, memEstimator)
+			m := NewHTTPHandler(log, ClientConfig{}, []string{"*"})
 			req := httptest.NewRequest(http.MethodOptions, "http://model-runner.docker.internal"+tt.path, http.NoBody)
 			req.Header.Set("Origin", "docker.com")
 			w := httptest.NewRecorder()

pkg/inference/memory/settings.go

@@ -15,7 +15,6 @@ import (
 	"github.com/docker/model-runner/pkg/distribution/distribution"
 	"github.com/docker/model-runner/pkg/distribution/registry"
 	"github.com/docker/model-runner/pkg/inference"
-	"github.com/docker/model-runner/pkg/inference/memory"
 	"github.com/docker/model-runner/pkg/internal/utils"
 	"github.com/docker/model-runner/pkg/logging"
 	"github.com/docker/model-runner/pkg/middleware"
@@ -38,8 +37,6 @@ type HTTPHandler struct {
 	httpHandler http.Handler
 	// lock is used to synchronize access to the models manager's router.
 	lock sync.RWMutex
-	// memoryEstimator is used to calculate runtime memory requirements for models.
-	memoryEstimator memory.MemoryEstimator
 	// manager handles business logic for model operations.
 	manager *Manager
 }
@@ -56,13 +53,12 @@ type ClientConfig struct {
 }
 
 // NewHTTPHandler creates a new model's handler.
-func NewHTTPHandler(log logging.Logger, c ClientConfig, allowedOrigins []string, memoryEstimator memory.MemoryEstimator) *HTTPHandler {
+func NewHTTPHandler(log logging.Logger, c ClientConfig, allowedOrigins []string) *HTTPHandler {
 	// Create the manager.
 	m := &HTTPHandler{
-		log:             log,
-		router:          http.NewServeMux(),
-		memoryEstimator: memoryEstimator,
-		manager:         NewManager(log.WithFields(logrus.Fields{"component": "service"}), c),
+		log:     log,
+		router:  http.NewServeMux(),
+		manager: NewManager(log.WithFields(logrus.Fields{"component": "service"}), c),
 	}
 
 	// Register routes.
@@ -163,23 +159,7 @@ func (h *HTTPHandler) handleCreateModel(w http.ResponseWriter, r *http.Request)
 	// Normalize the model name to add defaults
 	request.From = NormalizeModelName(request.From)
 
-	// Pull the model. In the future, we may support additional operations here
-	// besides pulling (such as model building).
-	if memory.RuntimeMemoryCheckEnabled() && !request.IgnoreRuntimeMemoryCheck {
-		h.log.Infof("Will estimate memory required for %q", request.From)
-		proceed, req, totalMem, err := h.memoryEstimator.HaveSufficientMemoryForModel(r.Context(), request.From, nil)
-		if err != nil {
-			h.log.Warnf("Failed to validate sufficient system memory for model %q: %s", request.From, err)
-			// Prefer staying functional in case of unexpected estimation errors.
-			proceed = true
-		}
-		if !proceed {
-			errstr := fmt.Sprintf("Runtime memory requirement for model %q exceeds total system memory: required %d RAM %d VRAM, system %d RAM %d VRAM", request.From, req.RAM, req.VRAM, totalMem.RAM, totalMem.VRAM)
-			h.log.Warnf(errstr)
-			http.Error(w, errstr, http.StatusInsufficientStorage)
-			return
-		}
-	}
+	// Pull the model
 	if err := h.manager.Pull(request.From, request.BearerToken, r, w); err != nil {
 		if errors.Is(err, context.Canceled) || errors.Is(err, context.DeadlineExceeded) {
 			h.log.Infof("Request canceled/timed out while pulling model %q", request.From)