Changeset 201 for trunk

Timestamp:

10/16/09 11:47:08 (15 years ago)

Author:

bas

Message:

examples/new_rack_pbsmon.py:

• use new data structure + automatically determine the size of the cluster:
  • number of racks
  • number of node per rack

Location:

trunk

Files:

• debian/changelog (modified) (1 diff)
• examples/new_rack_pbsmon.py (modified) (6 diffs)

trunk/debian/changelog

@@ -5 +5 @@
         structure.
     Author: Bas van der Vlies
+
+ * new_rack_pbsmon.py
+        Rewrite to new data structure and automatically determine how many nodes
+        and racks a cluster has. 
+   Author: Bas van der Vlies
 
  -- Bas van der Vlies <bas@rc.sara.nl>  Mon, 18 May 2009 17:03:16 +0200

trunk/examples/new_rack_pbsmon.py

@@ -15 +15 @@
 import string
 import getopt
+import re
 
 import pbs
 from PBSQuery import PBSQuery
 from PBSQuery import PBSError
+
+# A node has the following syntax gb-r10n10
+#  r10 is rack name -> skip one char --> gives us rack number = 10
+#  n10 is node name -> skip one char --> gives us node number = 10
+# Then we have to set these variables to determine automatically the 
+# number of nodes and racks 
+#
+NODE_EXPR = re.compile(r"""
+        (?P<racknr>r[0-9]+)
+        (?P<nodenr>n[0-9]+)
+""", re.VERBOSE) 
+
+SKIP_CHARS_RACK = 1 
+SKIP_CHARS_NODE = 1 
+
+NODE_SPLIT_STR = '-'
+NODE_STR = 'r%dn%d'
+
+
+
 
 
@@ -54 +75 @@
 
 
+def determine_limits(nodes):
+        """
+        Determiine the the limits of the cluster:
+          - How many racks are in the cluster
+          - What is the max amount of nodes per rack
+        """
+        rack_max = node_max = 0
+        for id in nodes:
+                result = NODE_EXPR.search(id)
+                if result:
+
+                        number = int(result.group('racknr')[SKIP_CHARS_RACK:])
+                        if number > rack_max:
+                                rack_max = number
+
+                        number = int(result.group('nodenr')[SKIP_CHARS_NODE:])
+                        if number > node_max:
+                                node_max = number
+
+        return (rack_max+1, node_max+1)
+                
+
 def pbsmon(server = None):
         global NODES_PER_RACK, N_RACKS, PBS_STATES, OPT_WIDE
@@ -79 +122 @@
         node_dict = {}
 
+        number_of_racks, nodes_per_rack = determine_limits(nodes)
+
         for id in nodes:
 
@@ -93 +138 @@
                 state_char = PBS_STATES[state]
 
-#               print 'TD: ', nodes[id].name, nodes[id].is_free() ,nodes[id].has_job()
+                #print 'TD: ', nodes[id].name, nodes[id].is_free() ,nodes[id].has_job()
 
                 if nodes[id].is_free() and nodes[id].has_job():         # single job
-#                       print 'TD: %s' % nodename, node
+                        #print 'TD: %s' % id, nodes[id]
                         state_char = PBS_STATES[pbs_ND_single]
 
-#               print 'TD: %s %s' % (nodename, state_char)
-                dummy = string.split(nodes[id].name, '-')
+                #print 'TD: %s %s' % (nodename, state_char)
+
+                # Remove the prefix of the node, eg gb-r15n1 --> r15n1
+                #
+                dummy = string.split(nodes[id].name, NODE_SPLIT_STR)
                 node_dict[dummy[1]] = state_char
+
+
 
 # print header lines
         print '  ',
-        for rack in xrange(START_RACK, N_RACKS+1):
+        for rack in xrange(START_RACK, number_of_racks):
                 if not (rack % 10):
                         print '%d' % (rack / 10),
@@ -116 +166 @@
 
         print '  ',
-        for rack in xrange(START_RACK, N_RACKS+1):
+        for rack in xrange(START_RACK, number_of_racks):
                 print '%d' % (rack % 10),
                 if OPT_WIDE:
@@ -123 +173 @@
 
 # print nodes with r%dn%d naming scheme
-        for node_nr in xrange(1, NODES_PER_RACK+1):
+        for node_nr in xrange(1, nodes_per_rack):
                 print '%2d' % node_nr,
 
-                for rack in xrange(START_RACK, N_RACKS+1):
-                        node_name = 'r%dn%d' % (rack, node_nr)
+                for rack in xrange(START_RACK, number_of_racks):
+                        node_name = NODE_STR %(rack, node_nr)
 
                         if node_dict.has_key(node_name):