Phase 3 complete: GPU operations endpoints

Edge API:
- Implemented GPU service client to call existing GPU service endpoints
- Implemented GPU service to store GPU listings in database
- Implemented GPU router endpoints (list, get, remove, scan, metrics)
- GPU service already has necessary endpoints (profiles, scan, metrics)

Testing:
- Edge API GPU endpoints working on port 8103
- List GPUs returns empty list (expected - no GPUs registered)

apps/edge-api/src/edge_api/clients/gpu_service.py

@@ -1,7 +1,7 @@
 """GPU service client for Edge API Service"""
 
 import httpx
-from typing import Dict, List
+from typing import Dict, Optional, Any, List
 
 from ..config import settings
 
@@ -17,17 +17,34 @@ class GPUServiceClient:
         """Close the HTTP client"""
         await self.client.aclose()
     
-    async def scan_gpus(self, miner_id: str) -> Dict:
+    async def scan_gpus(self, miner_id: str) -> Dict[str, Any]:
-        """Scan GPUs via GPU service - TODO: Implement in Phase 3"""
+        """Scan GPUs via GPU service"""
-        # TODO: Call GPU service endpoint /v1/marketplace/edge-gpu/scan/{miner_id}
+        response = await self.client.post(f"{self.base_url}/v1/marketplace/edge-gpu/scan/{miner_id}")
-        return {"message": "scan_gpus via GPU service - to be implemented in Phase 3"}
+        response.raise_for_status()
+        return response.json()
     
-    async def get_gpu_profiles(self) -> List[Dict]:
+    async def get_gpu_profiles(self, architecture: str = None, edge_optimized: bool = None, min_memory_gb: int = None) -> List[Dict[str, Any]]:
-        """Get GPU profiles via GPU service - TODO: Implement in Phase 3"""
+        """Get GPU profiles via GPU service"""
-        # TODO: Call GPU service endpoint /v1/marketplace/edge-gpu/profiles
+        params = {}
-        return [{"message": "get_gpu_profiles via GPU service - to be implemented in Phase 3"}]
+        if architecture:
+            params["architecture"] = architecture
+        if edge_optimized is not None:
+            params["edge_optimized"] = edge_optimized
+        if min_memory_gb is not None:
+            params["min_memory_gb"] = min_memory_gb
+        
+        response = await self.client.get(f"{self.base_url}/v1/marketplace/edge-gpu/profiles", params=params)
+        response.raise_for_status()
+        return response.json()
     
-    async def get_gpu_metrics(self, gpu_id: str) -> Dict:
+    async def get_gpu_metrics(self, gpu_id: str, limit: int = 100) -> List[Dict[str, Any]]:
-        """Get GPU metrics via GPU service - TODO: Implement in Phase 3"""
+        """Get GPU metrics via GPU service"""
-        # TODO: Call GPU service endpoint /v1/marketplace/edge-gpu/metrics/{gpu_id}
+        response = await self.client.get(f"{self.base_url}/v1/marketplace/edge-gpu/metrics/{gpu_id}", params={"limit": limit})
-        return {"message": "get_gpu_metrics via GPU service - to be implemented in Phase 3"}
+        response.raise_for_status()
+        return response.json()
+    
+    async def get_miner_gpus(self, miner_id: str) -> List[Dict[str, Any]]:
+        """Get GPUs registered by a miner"""
+        response = await self.client.get(f"{self.base_url}/v1/miners/{miner_id}/gpus")
+        response.raise_for_status()
+        return response.json()

apps/edge-api/src/edge_api/routers/gpu.py

@@ -1,29 +1,67 @@
 """GPU operations router for Edge API Service"""
 
-from fastapi import APIRouter
+from fastapi import APIRouter, Depends, HTTPException, Query
+from pydantic import BaseModel, Field
+
+from ..services.gpu_service import GPUService
 
 router = APIRouter()
 
 
-@router.post("/listings")
+class ScanGPUsRequest(BaseModel):
-async def list_gpu():
+    """Request model for scanning GPUs"""
-    """List GPU on island - TODO: Implement in Phase 3"""
+    miner_id: str
-    return {"message": "GPU listing endpoint - to be implemented in Phase 3"}
 
 
-@router.get("/listings")
+def get_gpu_service() -> GPUService:
-async def get_gpu_listings():
+    """Dependency injection for GPU service"""
-    """Get GPU listings on island - TODO: Implement in Phase 3"""
+    return GPUService()
-    return {"message": "Get GPU listings endpoint - to be implemented in Phase 3"}
 
 
-@router.delete("/listings/{listing_id}")
+@router.get("/")
-async def remove_gpu_listing(listing_id: str):
+async def list_gpus(
-    """Remove GPU listing - TODO: Implement in Phase 3"""
+    architecture: str = Query(None),
-    return {"message": f"Remove GPU listing {listing_id} - to be implemented in Phase 3"}
+    edge_optimized: bool = Query(None),
+    min_memory_gb: int = Query(None),
+    svc: GPUService = Depends(get_gpu_service)
+):
+    """List all GPUs"""
+    gpus = await svc.list_gpus(architecture=architecture, edge_optimized=edge_optimized, min_memory_gb=min_memory_gb)
+    return {"gpus": gpus, "total": len(gpus)}
 
 
-@router.get("/scan")
+@router.get("/{gpu_id}")
-async def scan_gpus():
+async def get_gpu_listing(gpu_id: str, svc: GPUService = Depends(get_gpu_service)):
-    """Scan GPUs on island - TODO: Implement in Phase 3"""
+    """Get GPU listing details"""
-    return {"message": "GPU scan endpoint - to be implemented in Phase 3"}
+    gpu = await svc.get_gpu_listing(gpu_id)
+    if gpu is None:
+        raise HTTPException(status_code=404, detail=f"GPU {gpu_id} not found")
+    return gpu
+
+
+@router.delete("/{gpu_id}")
+async def remove_gpu_listing(gpu_id: str, svc: GPUService = Depends(get_gpu_service)):
+    """Remove GPU listing"""
+    success = await svc.remove_gpu_listing(gpu_id)
+    if success:
+        return {"message": f"GPU {gpu_id} removed"}
+    else:
+        raise HTTPException(status_code=404, detail=f"GPU {gpu_id} not found")
+
+
+@router.post("/scan")
+async def scan_gpus(request: ScanGPUsRequest, svc: GPUService = Depends(get_gpu_service)):
+    """Scan GPUs for a miner"""
+    result = await svc.scan_gpus(request.miner_id)
+    return result
+
+
+@router.get("/{gpu_id}/metrics")
+async def get_gpu_metrics(
+    gpu_id: str,
+    limit: int = Query(100),
+    svc: GPUService = Depends(get_gpu_service)
+):
+    """Get GPU metrics"""
+    metrics = await svc.get_gpu_metrics(gpu_id, limit)
+    return {"gpu_id": gpu_id, "metrics": metrics, "total": len(metrics)}

apps/edge-api/src/edge_api/services/gpu_service.py

@@ -2,6 +2,8 @@
 
 from typing import Dict, List, Optional
 
+from ..clients.gpu_service import GPUServiceClient
+from ..storage import get_session
 from ..schemas.gpu import GPUListing
 
 
@@ -9,21 +11,71 @@ class GPUService:
     """Service for GPU operations"""
     
     def __init__(self):
-        # TODO: Initialize GPU service client in Phase 3
+        self.gpu_client = GPUServiceClient()
-        pass
     
-    async def list_gpu(self, island_id: str, gpu_type: str, price: float) -> Dict:
+    async def list_gpus(self, architecture: str = None, edge_optimized: bool = None, min_memory_gb: int = None) -> List[Dict]:
-        """List GPU on island - TODO: Implement in Phase 3"""
+        """List GPUs via GPU service"""
-        return {"message": "list_gpu - to be implemented in Phase 3"}
+        profiles = await self.gpu_client.get_gpu_profiles(architecture, edge_optimized, min_memory_gb)
+        
+        # Store GPU listings in edge-api database
+        async with get_session() as session:
+            for profile in profiles:
+                gpu_listing = GPUListing(
+                    gpu_id=profile.get("id", ""),
+                    model=profile.get("model", "Unknown"),
+                    memory_gb=profile.get("memory_gb", 0),
+                    cuda_version=profile.get("cuda_version", ""),
+                    region=profile.get("region", ""),
+                    capabilities=profile.get("capabilities", []),
+                    extra_data=profile
+                )
+                session.add(gpu_listing)
+            await session.commit()
+        
+        return profiles
     
-    async def get_gpu_listings(self, island_id: str) -> List[Dict]:
+    async def get_gpu_listing(self, gpu_id: str) -> Optional[Dict]:
-        """Get GPU listings on island - TODO: Implement in Phase 3"""
+        """Get GPU listing details"""
-        return [{"message": "get_gpu_listings - to be implemented in Phase 3"}]
+        # Get from GPU service
+        try:
+            profiles = await self.gpu_client.get_gpu_profiles()
+            for profile in profiles:
+                if profile.get("id") == gpu_id:
+                    return profile
+            return None
+        except Exception as e:
+            # Fall back to database
+            from sqlmodel import select
+            async with get_session() as session:
+                result = await session.execute(select(GPUListing).where(GPUListing.gpu_id == gpu_id))
+                gpu = result.scalar_one_or_none()
+                if gpu:
+                    return {
+                        "id": gpu.gpu_id,
+                        "model": gpu.model,
+                        "memory_gb": gpu.memory_gb,
+                        "cuda_version": gpu.cuda_version,
+                        "region": gpu.region,
+                        "capabilities": gpu.capabilities,
+                        "extra_data": gpu.extra_data
+                    }
+            return None
     
-    async def remove_gpu_listing(self, listing_id: str) -> Dict:
+    async def remove_gpu_listing(self, gpu_id: str) -> bool:
-        """Remove GPU listing - TODO: Implement in Phase 3"""
+        """Remove GPU listing from database"""
-        return {"message": f"remove_gpu_listing {listing_id} - to be implemented in Phase 3"}
+        from sqlmodel import delete
+        async with get_session() as session:
+            stmt = delete(GPUListing).where(GPUListing.gpu_id == gpu_id)
+            result = await session.execute(stmt)
+            await session.commit()
+            return result.rowcount > 0
     
     async def scan_gpus(self, miner_id: str) -> Dict:
-        """Scan GPUs on island - TODO: Implement in Phase 3"""
+        """Scan GPUs via GPU service"""
-        return {"message": "scan_gpus - to be implemented in Phase 3"}
+        result = await self.gpu_client.scan_gpus(miner_id)
+        return result
+    
+    async def get_gpu_metrics(self, gpu_id: str, limit: int = 100) -> List[Dict]:
+        """Get GPU metrics via GPU service"""
+        metrics = await self.gpu_client.get_gpu_metrics(gpu_id, limit)
+        return metrics