The UCP Advisor Service Console VM should be deployed on one or more management nodes. The UCP Advisor VM needs a direct connection to all of the converged infrastructure management and Fibre Channel networks.
The UCP Advisor software must be deployed on the Master Node VM. The management node requires a dual connection to ensure full connectivity into the UCP infrastructure. The following figure shows the cluster configuration:
The different modes of deploying the management nodes are:
- Customer-supplied node (already existing management node in which UCP Advisor and the necessary components can be installed).
- Customer uses a Hitachi-supplied DS120 management node.
- Optional configuration where the UCP Advisor Gateway VM is installed on the compute node, while the UCP Advisor Master Node VM resides on the customer management node.
The management node DS120 front view is shown in the following figure:
UCP Advisor should be deployed on VMware high-availability clustered servers and may coexist with non-mission-critical applications. UCP Advisor should not manage the servers that it is deployed on, with the exception of the UCP Advisor Gateway VM.
The following table lists the internal structure of the management node:
Resource | Quantity |
---|---|
RAM | 8x32 GB DDR4 R-DIMM (256 GB) per node |
CPUs (2 CPUs per node) | Intel Xeon® E5-2680 v4, 14c 2.3 GHz (Skylake configuration) |
Local storage | 2 TB |
Emulex Lpe31002 16/32 GB (Gen6 FC) PCIe | |
2-port 10 GBE (Intel) OCP Mezzanine card | |
1 GB dedicated management port |