Table 9 Hadoop configurations.
From: Exploring the effects and potential of unlocked I/O-powered single board computer clusters
mapred-site.xml | Value |
|---|---|
yarn.app.mapreduce.am.resource.mb | 1536 (default) |
yarn.app.mapreduce.am.resource.cpu-vcores | 1 (default) |
mapreduce.map.memory.mb | 1024 (default) |
mapreduce.reduce.memory.mb | 1024 (default) |
mapreduce.map.cpu.vcores | 1 (default) |
mapreduce.reduce.cpu.vcores | 1 (default) |
mapreduce.job.reduces | 2 |
hdfs-site.xml | Value |
|---|---|
dfs.replication | 1 |
dfs.namenode.name.dir | /hdfs/namenode (microSD) or |
/mnt/nvme/hdfs/namenode (SSD) | |
dfs.namenode.data.dir | /hdfs/datanode (microSD) or |
/mnt/nvme/hdfs/datanode (SSD) |
yarn-site.xml | Value |
|---|---|
yarn.nodemanager.resource.memory-mb | 8192 (default) |
yarn.nodemanager.resource.cpu-vcores | 8 (default) |
yarn.scheduler.minimum-allocation-mb | 1024 (default) |
yarn.scheduler.maximum-allocation-mb | 4096 (default) |
yarn.scheduler.minimum-allocation-vcores | 1 (default) |
yarn.scheduler.maximum-allocation-vcores | 32 (default) |
yarn.nodemanager.vmem-pmem-ratio | 2.1 (default) |
