bin/generate_cluster_xml.py

   1 #!/usr/bin/python
   2
   3 #
   4 #   Copyright 2013 LinkedIn, Inc
   5 #
   6 #  Licensed under the Apache License, Version 2.0 (the "License");
   7 #   you may not use this file except in compliance with the License.
   8 #   You may obtain a copy of the License at
   9 #
  10 #      http://www.apache.org/licenses/LICENSE-2.0
  11 #
  12 #  Unless required by applicable law or agreed to in writing, software
  13 #  distributed under the License is distributed on an "AS IS" BASIS,
  14 #  WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
  15 #  See the License for the specific language governing permissions and
  16 #  limitations under the License.
  17
  18 #  Python 2.7+ required
  19 #  This script encapsulates the cluster.xml generation for a zoned and a non-zoned
  20 #  cluster. Passing --zones <num of zones> switch to the script generates a zoned cluster
  21 #  config. A non zoned cluster is generated otherwise.
  22 #
  23 #  The newly generated cluster.xml file is placed in the output dir.
  24 #
  25 #  Example use for a zoned cluster :
  26 #  python generate_cluster_xml.py --file <file with host names, one host per line>
  27 #                                 --name <name of the cluster>
  28 #                                 --nodes <number of nodes>
  29 #                                 --partitions <number of partitions>
  30 #                                 --sock-port <port no>
  31 #                                 --admin-port <port no>
  32 #                                 --http-port <port no>
  33 #                                 --current-stores <current_stores.xml>
  34 #                                 --output-dir <output directory>
  35 #                                 --zones <number of zones>
  36 #
  37 #  For non zoned cluster use :
  38 #  python generate_cluster_xml.py --file <file with host names, one host per line>
  39 #                                 --name <name of the cluster>
  40 #                                 --nodes <number of nodes>
  41 #                                 --partitions <number of partitions>
  42 #                                 --sock-port <port no>
  43 #                                 --admin-port <port no>
  44 #                                 --http-port <port no>
  45 #                                 --current-stores <current_stores.xml>
  46 #                                 --output-dir <output directory>
  47 #
  48 # Note the absence of the --zones switch for the non zoned cluster use case.
  49
  50 import sys
  51 import random
  52 import os
  53 import errno
  54 import subprocess
  55
  56 try:
  57     import argparse
  58 except ImportError:
  59     print "Python 2.7 or higher is needed"
  60
  61
  62 # Get a random seed
  63 rseed = int(random.randint(00000000001,99999999999))
  64
  65 # Setup and argument parser
  66 parser = argparse.ArgumentParser(description='Build a voldemort cluster.xml.')
  67 # Add supported arguments
  68 parser.add_argument('-f', '--file', type=str, dest='file',
  69                     help='the file of the list of hosts(one per line)')
  70 parser.add_argument('-N', '--name', type=str, default='voldemort', dest='name',
  71                     help='the name you want to give the cluster')
  72 parser.add_argument('-n', '--nodes', type=int, default=6, dest='nodes',
  73                     help='the number of nodes in the cluster')
  74 parser.add_argument('-p', '--partitions', type=int, default=1500,
  75                     dest='partitions', help='number of partitions')
  76 parser.add_argument('-sp', '--socket-port', type=int, default=6666,
  77                     dest='sock_port', help='socket port number')
  78 parser.add_argument('-ap', '--admin-port', type=int, default=6667,
  79                     dest='admin_port', help='admin port number')
  80 parser.add_argument('-hp', '--http-port', type=int, default=6665,
  81                     dest='http_port', help='http port number')
  82 parser.add_argument('-s', '--current-stores', type=str, default= "config/tools/dummy-stores-3zoned.xml",
  83                     dest='current_stores',
  84                     help='Path to current stores xml. If you do not have info about the stores yet'
  85                          'use config/tools/dummy-stores.xml from the root voldemort home folder.')
  86 parser.add_argument('-o', '--output-dir', type=str, dest='output_dir',
  87                     help='output directory location')
  88 parser.add_argument('-z', '--zones', type=int, dest='zones',
  89                     help='For non zoned clusters do not provide this argument.'
  90                          'For zoned clusters provide this argument with at least two zones.')
  91
  92 genType = parser.add_mutually_exclusive_group()
  93 genType.add_argument('--seed', type=int, default=rseed, dest='seed',
  94                     help='seed for randomizing partition distribution')
  95
  96 # Parse arguments
  97 args = parser.parse_args()
  98
  99 # Check if the input file exists
 100 try:
 101    with open(args.file): pass
 102 except IOError:
 103    print 'File does not exist'
 104
 105 # create output-dir if it does not exist
 106 try:
 107     os.makedirs(args.output_dir)
 108 except OSError as exception:
 109     if exception.errno != errno.EEXIST:
 110         raise
 111
 112 # Open a new file named cluster.xml
 113 clusterXMLFilePath = os.path.join(os.path.abspath(args.output_dir), 'cluster.xml')
 114 fileHandle = open(clusterXMLFilePath, 'w')
 115
 116 # TODO : It would be ideal to have the script accept a list of zone ids.
 117 if args.zones:
 118   zones = args.zones
 119   if (zones == 1):
 120       print "For non zoned clusters do not provide this argument."
 121       print "For zoned clusters provide this argument with at least two zones."
 122       sys.exit(1)
 123   if (args.nodes % zones) != 0:
 124     print "Number of nodes must be evenly divisible by number of zones"
 125     sys.exit(1)
 126
 127 # Store arguments
 128 if args.file:
 129   hostList = open(args.file).readlines()
 130   nodes = len(hostList)
 131 else:
 132   nodes = args.nodes
 133 partitions = args.partitions
 134 name = args.name
 135 http_port = args.http_port
 136 sock_port = args.sock_port
 137 admin_port = args.admin_port
 138 seed = args.seed
 139 vold_home = os.pardir
 140 current_stores = os.path.join(vold_home, args.current_stores);
 141
 142 # Generate the full list of partition IDs
 143 part_ids = range(partitions)
 144 if part_ids < 1500:
 145   print 'Warning : The number of partitions seems to be low. Assuming max of 3 zones and 50 nodes ' \
 146         'per zone, a partition value of 1500 is recommended as it ensures an average of 10 ' \
 147         'partitions per node.'
 148   print 'Warning : The number of partitions seems to be low. Recommended value is 1500 or more.'
 149
 150 # Generate full list of zone IDs
 151 if args.zones:
 152   zone_ids = range(zones)
 153   zone_id = 0
 154
 155 # Shuffle up the partitions
 156 random.seed(seed)
 157 random.shuffle(part_ids)
 158
 159 # Printing cluster.xml
 160 print >> fileHandle, "<cluster>"
 161 print >> fileHandle, "  <name>%s</name>" % name
 162
 163 if args.zones:
 164   for i in range(args.zones):
 165     print >> fileHandle, "  <zone>"
 166     print >> fileHandle, "    <zone-id>%d</zone-id>" % i
 167     proximityList = list()
 168     for j in range(1, len(zone_ids) ):
 169       proximityList.append(zone_ids[(i+j)%len(zone_ids)])
 170     print >> fileHandle, "    <proximity-list>%s</proximity-list>" % str(proximityList).strip('[]')
 171     print >> fileHandle, "  </zone>"
 172
 173 # TODO : Currently, random partitions are assigned to the nodes in a round robin fashion.
 174 # A better approach would be to have some intelligence in the allocation such that
 175 # consecutive partition-ids do not land on the same node.
 176
 177 for i in xrange(nodes):
 178   j = i
 179   node_partitions = list()
 180   while j < len(part_ids):
 181     node_partitions.append(str(part_ids[j]))
 182     j += nodes;
 183   partitionslist = ", ".join(node_partitions);
 184
 185   print >> fileHandle, "  <server>"
 186   print >> fileHandle, "    <id>%d</id>" % i
 187   if args.file:
 188     print >> fileHandle, "    <host>%s</host>" % hostList[i].strip()
 189   else:
 190     print >> fileHandle, "    <host>host%d</host>" % i
 191   print >> fileHandle, "    <http-port>%d</http-port>" % http_port
 192   print >> fileHandle, "    <socket-port>%d</socket-port>" % sock_port
 193   print >> fileHandle, "    <admin-port>%d</admin-port>" % admin_port
 194   print >> fileHandle, "    <partitions>%s</partitions>" % partitionslist
 195   # If zones are being used, assign a zone-id
 196   if args.zones:
 197     print >> fileHandle, "    <zone-id>%d</zone-id>" % zone_id
 198     if zone_id == (zones - 1):
 199       zone_id = 0
 200     else:
 201       zone_id += 1
 202   print >> fileHandle, "  </server>"
 203 print >> fileHandle, "</cluster>"
 204
 205 fileHandle.close()
 206
 207 # For zoned clusters call rebalance-new-cluster.sh
 208 if args.zones:
 209     scriptPath = vold_home + '/bin/rebalance-new-zoned-cluster.sh'
 210     cmd = [scriptPath, '-v', vold_home, '-c', clusterXMLFilePath, '-s', current_stores,
 211                    '-o', os.path.abspath(args.output_dir)]
 212     subprocess.call(cmd)