Setting up dedicated switches to connect to iSCSI arrays and connecting top of rack (ToR) switches, which will support both data and storage traffic, to those dedicated switches is a best practice for supporting the iSCSI requirements. This provides isolation and security for the iSCSI traffic and reduces the risk of congestion or latency on the storage network1.

IMET routes advertise the MAC addresses that the VTEP has learned locally in a VXLAN is not a correct description of routing information advertised by a VXLAN Tunnel Endpoint (VTEP) that uses EVPN. IMET routes are one of the types of routes that EVPN uses to advertise multicast information for VXLAN networks. IMET routes advertise the IP addresses of VTEPs that can join multicast groups for VXLAN segments2.

The show dcbx interface command shows the current DCBx status and the configuration of PFC, ETS, and application priority applied on the interface and the status of the TLVs received from the peer1. The Application Priority Map section shows the protocol, port/type and priority for both local and remote advertisements. Therefore, this is where you can see whether the connected converged network adapter (CNA) has accepted the application priorities advertised with DCBX. Reference: https://www.arubanetworks.com/techdocs/AOS-CX/AOSCX-CLI-Bank/cli_8400/Content/Chp_DCBx/DCBx_cmds/sho-dcb-int.htm

Aruba Fabric Composer (AFC) with HPE StoreServ Management Console (SSMC) integration is possible. AFC is a software-defined networking solution that simplifies the management and orchestration of data center networks1. It can integrate with various data center software, such as VMware, Ansible, and Kubernetes1. SSMC is a web-based management tool that provides a unified interface for managing HPE 3PAR StoreServ storage systems2. AFC can integrate with SSMC to discover and visualize the storage network infrastructure and provide end-to-end visibility and troubleshooting1.
https://www.arubanetworks.com/products/switches/core-and-data-center/fabric-composer/

Creating two ERPS instances for the ring and assigning different VLANs and different RPLs to each instance is part of a valid strategy for load sharing traffic across the links in an ERPS solution1. ERPS is a protocol that provides protection and recovery for Ethernet traffic in a ring topology1. It uses a RPL to block one of the links in the ring and prevent loops1. By creating two ERPS instances with different RPLs, you can use both links in the ring for different VLANs and achieve load sharing1.