feat(azure-ai+gpu-ui): per-tenant Azure AI config + GPU health panel

- Per-tenant Azure AI config stored in tenants.tenant_config JSONB
- GET/PUT /api/tenants/{id}/ai-config + POST .../test connection
- api_key never returned to frontend (has_api_key: bool pattern)
- azure_ai.py resolves creds from tenant config when ai_enabled=True
- ai_tasks.py loads tenant config and passes it to validate_thumbnail
- Admin GPU Status section: probe button + status badge + last-checked time
- Notifications: _BELL_CHANNELS filter (notification+alert only in bell)
- Tenants.tsx: per-row Azure AI Config modal with URL auto-parse helper
- Remove duplicate in-memory /gpu-probe endpoints (kept DB-backed /probe/gpu)

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
This commit is contained in:
2026-03-08 21:04:09 +01:00
parent 34f89cc225
commit 22c29d5655
11 changed files with 792 additions and 24 deletions
+9
View File
@@ -12,6 +12,15 @@ _DEFAULT_TENANT_CONFIG = {
"fallback_material": "SCHAEFFLER_059999_FailedMaterial",
"notifications_enabled": True,
"invoice_prefix": "INV",
# Azure AI validation (per-tenant)
"ai_enabled": False,
"ai_api_key": None, # stored but never returned to frontend
"ai_endpoint": None, # e.g. https://myinstance.openai.azure.com
"ai_deployment": "gpt-4o",
"ai_api_version": "2024-02-01",
"ai_max_tokens": 500,
"ai_temperature": 0.1,
"ai_validation_prompt": None, # None = use DEFAULT_VALIDATION_PROMPT from azure_ai.py
}
+124 -1
View File
@@ -1,11 +1,16 @@
import uuid
from fastapi import APIRouter, Depends, HTTPException, status
from sqlalchemy.ext.asyncio import AsyncSession
from sqlalchemy import update
from app.database import get_db
from app.utils.auth import require_admin
from app.domains.tenants.schemas import TenantCreate, TenantUpdate, TenantOut
from app.domains.tenants.schemas import (
TenantCreate, TenantUpdate, TenantOut,
TenantAIConfigUpdate, TenantAIConfigOut,
)
from app.domains.tenants import service
from app.domains.tenants.models import Tenant
router = APIRouter(prefix="/tenants", tags=["tenants"])
@@ -77,3 +82,121 @@ async def delete_tenant(
status_code=status.HTTP_409_CONFLICT,
detail="Tenant not found or still has users assigned",
)
# ---------------------------------------------------------------------------
# Per-tenant Azure AI configuration
# ---------------------------------------------------------------------------
def _tenant_ai_config_out(tenant: Tenant) -> TenantAIConfigOut:
"""Build TenantAIConfigOut from a Tenant ORM object (never exposes api_key)."""
cfg = tenant.tenant_config or {}
return TenantAIConfigOut(
ai_enabled=cfg.get("ai_enabled", False),
ai_endpoint=cfg.get("ai_endpoint"),
ai_deployment=cfg.get("ai_deployment", "gpt-4o"),
ai_api_version=cfg.get("ai_api_version", "2024-02-01"),
has_api_key=bool(cfg.get("ai_api_key")),
ai_max_tokens=cfg.get("ai_max_tokens", 500),
ai_temperature=cfg.get("ai_temperature", 0.1),
ai_validation_prompt=cfg.get("ai_validation_prompt"),
)
@router.get("/{tenant_id}/ai-config", response_model=TenantAIConfigOut)
async def get_tenant_ai_config(
tenant_id: uuid.UUID,
db: AsyncSession = Depends(get_db),
_: object = Depends(require_admin),
):
"""Return AI config for a tenant (without the raw api_key)."""
tenant = await service.get_tenant(db, tenant_id)
if not tenant:
raise HTTPException(status_code=status.HTTP_404_NOT_FOUND, detail="Tenant not found")
return _tenant_ai_config_out(tenant)
@router.put("/{tenant_id}/ai-config", response_model=TenantAIConfigOut)
async def update_tenant_ai_config(
tenant_id: uuid.UUID,
body: TenantAIConfigUpdate,
db: AsyncSession = Depends(get_db),
_: object = Depends(require_admin),
):
"""Merge AI configuration into tenant_config JSONB.
If ai_api_key is None in the request body, the existing key is preserved.
"""
tenant = await service.get_tenant(db, tenant_id)
if not tenant:
raise HTTPException(status_code=status.HTTP_404_NOT_FOUND, detail="Tenant not found")
# Build the merged config dict
current_cfg: dict = dict(tenant.tenant_config or {})
current_cfg["ai_enabled"] = body.ai_enabled
current_cfg["ai_endpoint"] = body.ai_endpoint
current_cfg["ai_deployment"] = body.ai_deployment
current_cfg["ai_api_version"] = body.ai_api_version
current_cfg["ai_max_tokens"] = body.ai_max_tokens
current_cfg["ai_temperature"] = body.ai_temperature
current_cfg["ai_validation_prompt"] = body.ai_validation_prompt
# Only overwrite the stored key if a new value was actually provided
if body.ai_api_key is not None:
current_cfg["ai_api_key"] = body.ai_api_key
# Use direct SQL UPDATE to reliably persist JSONB mutations
await db.execute(
update(Tenant)
.where(Tenant.id == tenant_id)
.values(tenant_config=current_cfg)
)
await db.commit()
await db.refresh(tenant)
return _tenant_ai_config_out(tenant)
@router.post("/{tenant_id}/ai-config/test")
async def test_tenant_ai_config(
tenant_id: uuid.UUID,
db: AsyncSession = Depends(get_db),
_: object = Depends(require_admin),
):
"""Send a minimal ping to Azure OpenAI using the tenant's stored credentials.
Returns {"ok": true} or {"ok": false, "error": "human readable message"}.
"""
tenant = await service.get_tenant(db, tenant_id)
if not tenant:
raise HTTPException(status_code=status.HTTP_404_NOT_FOUND, detail="Tenant not found")
cfg = tenant.tenant_config or {}
api_key = cfg.get("ai_api_key")
endpoint = cfg.get("ai_endpoint")
deployment = cfg.get("ai_deployment", "gpt-4o")
api_version = cfg.get("ai_api_version", "2024-02-01")
if not api_key:
return {"ok": False, "error": "No API key configured for this tenant"}
if not endpoint:
return {"ok": False, "error": "No endpoint configured for this tenant"}
import asyncio
def _ping() -> dict:
try:
from openai import AzureOpenAI
client = AzureOpenAI(
api_key=api_key,
azure_endpoint=endpoint,
api_version=api_version,
)
client.chat.completions.create(
model=deployment,
messages=[{"role": "user", "content": "ping"}],
max_tokens=5,
)
return {"ok": True}
except Exception as exc:
return {"ok": False, "error": str(exc)}
return await asyncio.to_thread(_ping)
+22
View File
@@ -24,3 +24,25 @@ class TenantOut(BaseModel):
created_at: datetime
model_config = {"from_attributes": True}
class TenantAIConfigUpdate(BaseModel):
ai_enabled: bool = False
ai_endpoint: str | None = None
ai_deployment: str = "gpt-4o"
ai_api_version: str = "2024-02-01"
ai_api_key: str | None = None # optional — don't require re-entry
ai_max_tokens: int = 500
ai_temperature: float = 0.1
ai_validation_prompt: str | None = None
class TenantAIConfigOut(BaseModel):
ai_enabled: bool
ai_endpoint: str | None
ai_deployment: str
ai_api_version: str
has_api_key: bool # True if ai_api_key is set — never return the key itself
ai_max_tokens: int
ai_temperature: float
ai_validation_prompt: str | None