Update app.py

app.py

@@ -5,7 +5,7 @@ Pings all Spaces in an organization on a schedule to prevent them from sleeping.
 
 import os
 import json
-import threading
+from concurrent.futures import ThreadPoolExecutor, as_completed
 from datetime import datetime, timezone
 from pathlib import Path
 
@@ -16,15 +16,17 @@ from apscheduler.schedulers.background import BackgroundScheduler
 from apscheduler.triggers.interval import IntervalTrigger
 
 # Configuration
-ORG_NAME = os.environ.get("ORG_NAME", "mcp-hackathon")
+ORG_NAME = os.environ.get("ORG_NAME", "MCP-1st-Birthday")
 PING_INTERVAL_HOURS = int(os.environ.get("PING_INTERVAL_HOURS", "6"))
 REQUEST_TIMEOUT = int(os.environ.get("REQUEST_TIMEOUT", "30"))
+PARALLEL_REQUESTS = int(os.environ.get("PARALLEL_REQUESTS", "10"))  # Ping 10 Spaces at once
+HF_TOKEN = os.environ.get("HF_TOKEN", None)  # Optional: needed for private Spaces
 LOG_FILE = Path("run_logs.json")
 MAX_LOG_ENTRIES = 100  # Keep last N runs
 
 # Global state
 scheduler = BackgroundScheduler()
-api = HfApi()
+api = HfApi(token=HF_TOKEN) if HF_TOKEN else HfApi()
 
 
 def load_logs() -> list:
@@ -47,13 +49,20 @@ def save_logs(logs: list):
 
 def ping_space(space_id: str) -> dict:
     """Ping a single Space and return the result."""
-    url = f"https://huggingface.co/spaces/{space_id}"
+    # Convert space_id (org/name) to the actual app URL
+    # e.g., "MCP-1st-Birthday/my-app" -> "mcp-1st-birthday-my-app.hf.space"
+    org, name = space_id.split("/")
+    app_url = f"https://{org.lower()}-{name.lower()}.hf.space"
+    hf_page_url = f"https://huggingface.co/spaces/{space_id}"
+    
+    # Try the app URL first (this actually wakes up the Space)
     try:
-        response = requests.get(url, timeout=REQUEST_TIMEOUT)
+        response = requests.get(app_url, timeout=REQUEST_TIMEOUT)
         return {
             "space_id": space_id,
             "status": "success",
             "status_code": response.status_code,
+            "url_pinged": app_url,
             "error": None
         }
     except requests.Timeout:
@@ -61,19 +70,32 @@ def ping_space(space_id: str) -> dict:
             "space_id": space_id,
             "status": "timeout",
             "status_code": None,
+            "url_pinged": app_url,
             "error": f"Request timed out after {REQUEST_TIMEOUT}s"
         }
     except Exception as e:
-        return {
-            "space_id": space_id,
-            "status": "error",
-            "status_code": None,
-            "error": str(e)
-        }
+        # Fallback: try the HF page URL
+        try:
+            response = requests.get(hf_page_url, timeout=REQUEST_TIMEOUT)
+            return {
+                "space_id": space_id,
+                "status": "success",
+                "status_code": response.status_code,
+                "url_pinged": hf_page_url,
+                "error": None
+            }
+        except Exception as e2:
+            return {
+                "space_id": space_id,
+                "status": "error",
+                "status_code": None,
+                "url_pinged": app_url,
+                "error": str(e)
+            }
 
 
 def run_ping_job(triggered_by: str = "scheduler") -> dict:
-    """Run the ping job for all Spaces in the org."""
+    """Run the ping job for all Spaces in the org using parallel requests."""
     start_time = datetime.now(timezone.utc)
     
     # Get all Spaces in the org
@@ -96,11 +118,18 @@ def run_ping_job(triggered_by: str = "scheduler") -> dict:
         save_logs(logs)
         return run_result
     
-    # Ping each Space
+    # Ping Spaces in parallel (10 at a time by default)
     results = []
-    for space in spaces:
-        result = ping_space(space.id)
-        results.append(result)
+    space_ids = [space.id for space in spaces]
+    
+    with ThreadPoolExecutor(max_workers=PARALLEL_REQUESTS) as executor:
+        # Submit all ping tasks
+        future_to_space = {executor.submit(ping_space, space_id): space_id for space_id in space_ids}
+        
+        # Collect results as they complete
+        for future in as_completed(future_to_space):
+            result = future.result()
+            results.append(result)
     
     end_time = datetime.now(timezone.utc)
     duration = (end_time - start_time).total_seconds()
@@ -211,12 +240,15 @@ def get_logs_display() -> str:
 def get_status() -> str:
     """Get current scheduler status."""
     next_run = scheduler.get_jobs()[0].next_run_time if scheduler.get_jobs() else None
+    token_status = "✅ Configured" if HF_TOKEN else "❌ Not set (only public Spaces will be listed)"
     
     return f"""## Space Keeper Status
 
 **Organization:** `{ORG_NAME}`
 **Ping Interval:** Every {PING_INTERVAL_HOURS} hours
+**Parallel Requests:** {PARALLEL_REQUESTS} Spaces at once
 **Request Timeout:** {REQUEST_TIMEOUT} seconds
+**HF Token:** {token_status}
 **Next Scheduled Run:** {next_run.strftime('%Y-%m-%d %H:%M:%S UTC') if next_run else 'Not scheduled'}
 
 ---