Added exception handling

PILOTVERSION

@@ -1 +1 @@
-3.9.2.32
+3.9.2.33

pilot/util/constants.py

@@ -28,7 +28,7 @@
 RELEASE = '3'   # released number should be fixed at 3 for Pilot 3
 VERSION = '9'   # version number is '1' for first release, '0' until then, increased for bigger updates
 REVISION = '2'  # revision number should be reset to '0' for every new version release, increased for small updates
-BUILD = '32'     # build number should be reset to '1' for every new development cycle
+BUILD = '33'     # build number should be reset to '1' for every new development cycle
 
 SUCCESS = 0
 FAILURE = 1

pilot/util/psutils.py

@@ -335,10 +335,16 @@ def find_lingering_processes(parent_pid: int) -> list:
         return []
 
     lingering_processes = []
-    parent_process = psutil.Process(parent_pid)
-    for child in parent_process.children(recursive=True):
-        if child.status() != psutil.STATUS_ZOMBIE:
-            lingering_processes.append(child.pid)
+    try:
+        parent_process = psutil.Process(parent_pid)
+        for child in parent_process.children(recursive=True):
+            try:
+                if child.status() != psutil.STATUS_ZOMBIE:
+                    lingering_processes.append(child.pid)
+            except (psutil.NoSuchProcess, psutil.AccessDenied, psutil.ZombieProcess) as e:
+                logger.warning(f"[harmless] failed to get status for child process {child.pid}: {e}")
+    except (psutil.NoSuchProcess, psutil.AccessDenied, psutil.ZombieProcess, psutil.FileNotFoundError) as e:
+        logger.warning(f"[harmless] failed to get parent process {parent_pid}: {e}")
 
     return lingering_processes