1 name "hadoop_aws_t1-micro"
2 description "Hadoop cluster conf. on the AWS EC2 t1.micro instances."
3 # t1.micro: 1 vCPU, 0.615 GiB, no storage (EBS only)
4 # http://docs.aws.amazon.com/AWSEC2/latest/UserGuide/concepts_micro_instances.html
6 #cookbook_versions "hadoop" => "= 0.2.0"
10 'dfs.replication' => '2', # default: 3 (recommended)
13 # e.g. 'core-site.xml' => {'k1' => 'v1', 'k2' => 'v2'},
15 'HADOOP_HEAPSIZE' => '500', # default: 1000 MB
16 #'HADOOP_NAMENODE_INIT_HEAPSIZE' => '2000' # default: 1000 MB (ver. 2.x)
17 #'HADOOP_NAMENODE_OPTS' => '"${HADOOP_NAMENODE_OPTS} -Xmx500m"',
18 #'HADOOP_SECONDARYNAMENODE_OPTS' => '"${HADOOP_SECONDARYNAMENODE_OPTS} -Xmx500m"',
19 #'HADOOP_DATANODE_OPTS' => '"${HADOOP_DATANODE_OPTS} "',
20 #'HADOOP_BALANCER_OPTS' => '"${HADOOP_BALANCER_OPTS} "',
21 #'HADOOP_JOBTRACKER_OPTS' => '"${HADOOP_JOBTRACKER_OPTS} -Xmx500m"',
23 'yarn-env.sh' => { # ver. 2.x
24 'YARN_HEAPSIZE' => '500' # default: 1000 MB
25 #'YARN_RESOURCEMANAGER_OPTS' => '"${YARN_RESOURCEMANAGER_OPTS} -Xmx500m"',
26 #'YARN_NODEMANAGER_OPTS' => '"${YARN_NODEMANAGER_OPTS} "',
28 'mapred-env.sh' => { # ver. 2.x
29 #'HADOOP_JOB_HISTORYSERVER_HEAPSIZE' => '500', # default: 1000 MB
30 #'HADOOP_JOB_HISTORYSERVER_OPTS' => '"${HADOOP_JOB_HISTORYSERVER_OPTS} "',
35 'dfs.block.size' => '134217728', # default: 67108864 (bytes)
36 #'dfs.blocksize' => '134217728', # default: 134217728 (bytes) (ver. 2.x)
37 #'dfs.datanode.du.reserved' => '0', # default: 0 (byte)
39 #'dfs.support.append' => 'true',
40 #'dfs.datanode.max.xcievers' => '4096', # default: 4096 (ver. 1.1 or later), 256 (older ver.)
41 #'dfs.datanode.max.transfer.threads' => '4096', # default: 4096 (ver. 2.x)
43 'yarn-site.xml' => { # ver. 2.x
44 #'yarn.nodemanager.resource.cpu-vcores' => '8', # default: 8
45 #'yarn.nodemanager.resource.memory-mb' => '8192', # default: 8192
47 'mapred-site.xml' => {
48 # map tasks:reduce tasks = 4:3 or 2:1 or ...
49 'mapred.tasktracker.map.tasks.maximum' => '2', # default: 2
50 'mapred.tasktracker.reduce.tasks.maximum' => '1', # default: 2
51 #'mapred.child.java.opts' => '-Xmx1000m', # default: -Xmx200m
52 'mapred.reduce.slowstart.completed.maps' => '0.80', # default: 0.05
53 #'mapreduce.job.reduce.slowstart.completedmaps' => '0.80', # default: 0.05 (ver. 2.x)
54 #'tasktracker.http.threads' => '80', # default: 40
55 #'mapreduce.tasktracker.http.threads' => '80', # default: 40 (ver. 2.x)
56 #'mapred.reduce.parallel.copies' => SQRT(nodes), # default: 5
57 #'mapreduce.reduce.shuffle.parallelcopies' => SQRT(nodes), # default: 5 (ver. 2.x)
58 'mapred.submit.replication' => '2', # default: 10
59 #'mapreduce.client.submit.file.replication' => '3', # default: 10 (ver. 2.x)
61 'capacity-scheduler.xml' => {}
69 'hadoop' => hadoop_attributes