Topograph model import for use with toposim

README.md

@@ -99,12 +99,31 @@ systemctl disable topograph.service
 systemctl daemon-reload
 ```
 
-#### Testing the Service
-To verify the service is running correctly, you can use the following commands:
+#### Verifying Health
+To verify the service is healthy, you can use the following command:
 
 ```bash
 curl http://localhost:49021/healthz
+```
+
+#### Using Toposim
+To test the service on a simulated cluster, first add the following line to `/etc/topograph/topograph-config.yaml` so that any topology requests are forwarded to toposim.
+```bash
+forward_service_url: dns:localhost:49025
+```
 
+Then run the topograph service with the following commands, choosing any of the available cluster models:
+```bash
+/usr/local/bin/topograph -c /etc/topograph/topograph-config.yaml -m /etc/topograph/tests/models/<cluster-model>.yaml
+```
+
+You must them start the toposim service as such, using the same cluster model:
+```bash
+/usr/local/bin/topograph -m ./tests/models/<cluster-model>.yaml
+```
+
+And you can verify the topology results via the following commands:
+```bash
 id=$(curl -s -X POST -H "Content-Type: application/json" -d '{"provider":{"name":"test"},"engine":{"name":"test"}}' http://localhost:49021/v1/generate)
 
 curl -s "http://localhost:49021/v1/topology?uid=$id"

cmd/topograph/main.go

@@ -27,15 +27,18 @@ import (
 	"k8s.io/klog/v2"
 
 	"github.com/NVIDIA/topograph/pkg/config"
+	"github.com/NVIDIA/topograph/pkg/models"
 	"github.com/NVIDIA/topograph/pkg/server"
 )
 
 var GitTag string
 
 func main() {
 	var c string
+	var m string
 	var version bool
 	flag.StringVar(&c, "c", "/etc/topograph/topograph-config.yaml", "config file")
+	flag.StringVar(&m, "m", "", "model file to use when using the test provider")
 	flag.BoolVar(&version, "version", false, "show the version")
 
 	klog.InitFlags(nil)
@@ -47,13 +50,13 @@ func main() {
 		os.Exit(0)
 	}
 
-	if err := mainInternal(c); err != nil {
+	if err := mainInternal(c, m); err != nil {
 		klog.Errorf(err.Error())
 		os.Exit(1)
 	}
 }
 
-func mainInternal(c string) error {
+func mainInternal(c string, m string) error {
 	cfg, err := config.NewFromFile(c)
 	if err != nil {
 		return err
@@ -63,10 +66,18 @@ func mainInternal(c string) error {
 		return err
 	}
 
+	var model *models.Model = nil
+	if m != "" {
+		model, err = models.NewModelFromFile(m)
+		if err != nil {
+			return err
+		}
+	}
+
 	ctx, cancel := context.WithCancel(context.Background())
 	defer cancel()
 
-	server.InitHttpServer(ctx, cfg)
+	server.InitHttpServer(ctx, cfg, model)
 
 	var g run.Group
 	// Signal handler

cmd/toposim/main.go

@@ -26,6 +26,7 @@ import (
 	"github.com/oklog/run"
 	"k8s.io/klog/v2"
 
+	"github.com/NVIDIA/topograph/pkg/models"
 	"github.com/NVIDIA/topograph/pkg/toposim"
 )
 
@@ -50,7 +51,7 @@ func mainInternal() error {
 		return fmt.Errorf("must specify topology model path and listening port")
 	}
 
-	model, err := toposim.NewModelFromFile(path)
+	model, err := models.NewModelFromFile(path)
 	if err != nil {
 		return err
 	}

pkg/engines/k8s/labeler_test.go

@@ -38,7 +38,7 @@ func (l *testLabeler) AddNodeLabels(_ context.Context, nodeName string, labels m
 }
 
 func TestApplyNodeLabels(t *testing.T) {
-	root, _ := translate.GetTreeTestSet(true)
+	root, _ := translate.GetTreeTestSet(nil, true)
 	labeler := &testLabeler{data: make(map[string]map[string]string)}
 	data := map[string]map[string]string{
 		"Node201": {"topology.kubernetes.io/network-level-1": "S2", "topology.kubernetes.io/network-level-2": "S1"},

pkg/factory/provider.go

@@ -22,6 +22,7 @@ import (
 	"net/http"
 
 	"github.com/NVIDIA/topograph/pkg/common"
+	"github.com/NVIDIA/topograph/pkg/models"
 	"github.com/NVIDIA/topograph/pkg/providers/aws"
 	"github.com/NVIDIA/topograph/pkg/providers/baremetal"
 	"github.com/NVIDIA/topograph/pkg/providers/cw"
@@ -48,7 +49,7 @@ func GetProvider(provider string) (common.Provider, *common.HTTPError) {
 	case common.ProviderBM:
 		prv, err = baremetal.GetProvider()
 	case common.ProviderTest:
-		prv = GetTestProvider()
+		prv, err = GetTestProvider()
 	default:
 		return nil, common.NewHTTPError(http.StatusBadRequest, fmt.Sprintf("unsupported provider %q", provider))
 	}
@@ -60,16 +61,18 @@ func GetProvider(provider string) (common.Provider, *common.HTTPError) {
 	return prv, nil
 }
 
+var TestModel *models.Model = nil
+
 type testProvider struct {
 	tree          *common.Vertex
 	instance2node map[string]string
 }
 
-func GetTestProvider() *testProvider {
+func GetTestProvider() (*testProvider, error) {
 	p := &testProvider{}
-	p.tree, p.instance2node = translate.GetTreeTestSet(false)
+	p.tree, p.instance2node = translate.GetTreeTestSet(TestModel, false)
 
-	return p
+	return p, nil
 }
 
 func (p *testProvider) GetCredentials(_ map[string]string) (interface{}, error) {

pkg/toposim/model.go → pkg/models/model.go

@@ -14,7 +14,7 @@
  * limitations under the License.
  */
 
-package toposim
+package models
 
 import (
 	"fmt"

pkg/toposim/model_test.go → pkg/models/model_test.go

@@ -14,7 +14,7 @@
  * limitations under the License.
  */
 
-package toposim
+package models
 
 import (
 	"testing"
@@ -23,7 +23,7 @@ import (
 )
 
 func TestNewModelFromFile(t *testing.T) {
-	cfg, err := NewModelFromFile("testdata/toposim.yaml")
+	cfg, err := NewModelFromFile("../../tests/models/medium-h100.yaml")
 	require.NoError(t, err)
 
 	expected := &Model{