Changeset 790 for branches

Timestamp:

04/05/13 10:35:10 (10 years ago)

Author:

ramonb

Message:

• catch PBSError for PBSQuery.getnodes() just like with DataGatherer?.getJobData()
• incidentally PBS server can be non responsive: should not crash jobmond

File:

• branches/0.4/jobmond/jobmond.py (modified) (3 diffs)

branches/0.4/jobmond/jobmond.py

@@ -869 +869 @@
         
             l        = ['state']
-       
-            # TODO catch PBSError: like pq.getJobData
-            for name, node in self.pq.getnodes().items():
+
+            nodelist = self.getNodeData()
+
+            for name, node in nodelist.items():
 
                 if ( node[ 'state' ].find( "down" ) != -1 ):
@@ -1550 +1551 @@
             pass
 
+    def getNodeData( self ):
+
+        nodedict = { }
+
+        try:
+            nodedict = self.pq.getnodes()
+
+        except PBSError, detail:
+
+            debug_msg( 10, "PBS server unavailable, skipping until next polling interval: " + str( detail ) )
+
+        return nodedict
+
     def getJobData( self ):
 
@@ -1563 +1577 @@
         except PBSError, detail:
 
-            debug_msg( 10, "Caught PBS unavailable, skipping until next polling interval: " + str( detail ) )
+            debug_msg( 10, "PBS server unavailable, skipping until next polling interval: " + str( detail ) )
             return None