Home > Storage > PowerScale (Isilon) > Industry Solutions and Verticals > Hadoop and Big Data > Dell PowerScale and Cloudera CDP Private Cloud Base Reference Architecture > Sizing recommendation for physical nodes
The following table provides size recommendations for the physical nodes.
Component | Configuration | Description | Quantity |
Master nodes: Two-socket with 6–10 cores/socket > 2 GHz; | 2U 2-socket nodes with at least 256 GB RAM | These nodes house the Master services and serve as the gateway or edge device that connects the rest of the customer’s network to the Hadoop cluster. | Three (for scaling up to 100 cluster nodes). |
Compute instances: Two-socket with 6–10 cores/socket > 2 GHz; 2 x OS disks, 8 SATA or SAS drives or 2x SSDs
| At least 8 SATA or SAS drives, or 2 SSD drives for intermediate storage. | These nodes house the YARN node managers, and other required services. | Results in the field show that 1:2 ratio of PowerScale nodes to compute nodes is reasonable for most use cases. If there are heavy Impala workloads, use s 1:1.5 ratio. If PowerScale has 5 nodes, have 8 compute nodes. |
Utility or edge nodes: Two-socket with 6 to 10 cores/socket > 2 GHz; | 2U 2-socket nodes with at least 256 GB RAM | These nodes house the management services and serve as the gateway or edge device that connects the rest of the customer’s network to the Hadoop cluster. | 1 each for the cluster. |