02-05-2019 12:19 AM
We installed a Cloudera cluster last year for testing purposes, it was composed of 4 VMs : 1 as the cloudera manager and 3 as cloudera nodes and it seems to be running properly the needed services: HBase, HDFS, YARN and ZooKeeper.
The resources of our test Cloudera cluster are the following:
- Cloudera manager node: 3 vCPU, 32GB RAM, 150 GB disk
- Cloudera nodes: 3 vCPU, 16GB RAM, 150 GB disk
Now we want to install a new Cloudera cluster in the production environment and it will process a bigger amount of data but we are not sure if is there any problem for running this operational cluster on VMs instead of physical servers. Is anybody running these services on VMs without suffering performance issues?
Many thanks in advance!