test(api): cover health metrics and auth flows

test/api-server_test.ts

@@ -0,0 +1,92 @@
+import { assertEquals } from 'jsr:@std/assert@^1.0.0';
+import { ApiServer } from '../ts/api/server.ts';
+
+Deno.test('ApiServer serves health metrics and authenticated model listings', async () => {
+  const port = 18100 + Math.floor(Math.random() * 1000);
+  const server = new ApiServer(
+    {
+      host: '127.0.0.1',
+      port,
+      apiKeys: ['valid-key'],
+      cors: false,
+      corsOrigins: [],
+    },
+    {
+      async getAllStatus() {
+        return new Map([
+          ['vllm-1', { running: true, health: 'healthy' }],
+        ]);
+      },
+      async getAllAvailableModels() {
+        return new Map([
+          ['meta-llama/Llama-3.1-8B-Instruct', [{ type: 'vllm' }]],
+        ]);
+      },
+    } as never,
+    {
+      async getAllModels() {
+        return [
+          {
+            id: 'meta-llama/Llama-3.1-8B-Instruct',
+            engine: 'vllm',
+            source: { repo: 'meta-llama/Llama-3.1-8B-Instruct' },
+            capabilities: { chat: true },
+            requirements: { minVramGb: 18 },
+          },
+        ];
+      },
+    } as never,
+    {} as never,
+    {
+      getStatus() {
+        return {
+          localNode: null,
+          nodes: [],
+          models: {},
+          desiredDeployments: [],
+        };
+      },
+    } as never,
+  );
+
+  (server as unknown as {
+    gpuDetector: { detectGpus: () => Promise<unknown[]> };
+  }).gpuDetector = {
+    async detectGpus() {
+      return [{ id: 'nvidia-0' }];
+    },
+  };
+
+  await server.start();
+
+  try {
+    const healthResponse = await fetch(`http://127.0.0.1:${port}/health`);
+    const healthBody = await healthResponse.json();
+    assertEquals(healthResponse.status, 200);
+    assertEquals(healthBody.status, 'ok');
+    assertEquals(healthBody.models, 1);
+
+    const metricsResponse = await fetch(`http://127.0.0.1:${port}/metrics`);
+    const metricsBody = await metricsResponse.text();
+    assertEquals(metricsResponse.status, 200);
+    assertEquals(metricsBody.includes('modelgrid_uptime_seconds'), true);
+    assertEquals(metricsBody.includes('modelgrid_models_available 1'), true);
+
+    const unauthenticatedModels = await fetch(`http://127.0.0.1:${port}/v1/models`);
+    const unauthenticatedBody = await unauthenticatedModels.json();
+    assertEquals(unauthenticatedModels.status, 401);
+    assertEquals(unauthenticatedBody.error.type, 'authentication_error');
+
+    const authenticatedModels = await fetch(`http://127.0.0.1:${port}/v1/models`, {
+      headers: {
+        Authorization: 'Bearer valid-key',
+      },
+    });
+    const authenticatedBody = await authenticatedModels.json();
+    assertEquals(authenticatedModels.status, 200);
+    assertEquals(authenticatedBody.object, 'list');
+    assertEquals(authenticatedBody.data[0].id, 'meta-llama/Llama-3.1-8B-Instruct');
+  } finally {
+    await server.stop();
+  }
+});