Don't stop reporting/destroying, even if something bad happens to an instance.

slapos/grid/slapgrid.py

@@ -809,6 +809,7 @@ class Slapgrid(object):
     #We loop on the different computer partitions
     computer_partition_list = slap_computer_usage.getComputerPartitionList()
     for computer_partition in computer_partition_list:
+      try:
         computer_partition_id = computer_partition.getId()
         
         #We want execute all the script in the report folder
@@ -829,7 +830,6 @@ class Slapgrid(object):
         
         failed_script_list = []
         for script in script_list_to_run:
-
           invocation_list = []
           invocation_list.append(os.path.join(instance_path, 'etc', 'report',
             script))
@@ -865,6 +865,13 @@ class Slapgrid(object):
               (invocation_list, result))
           if len(failed_script_list):
             computer_partition.error('\n'.join(failed_script_list))
+      # Whatever happens, don't stop processing other instances
+      except Exception:
+        computer_partition_id = computer_partition.getId()
+        exception = traceback.format_exc()
+        issue = "Cannot run usage script(s) for %r: %s" % (
+            computer_partition_id, exception)
+        logger.info(issue)
 
     #Now we loop through the different computer partitions to report
     report_usage_issue_cp_list = []