Rack-mounted SAN and DAS arrays typically expose a management controller with its own SNMP agent, separate from any host that consumes the storage.
SAN arrays usually have a dedicated Ethernet management port distinct from their Fibre Channel or iSCSI data ports; that's where SNMP is configured, not on the data-path interfaces.
The exact menu varies by vendor and generation, but the pattern is consistent with other enterprise storage: a community string or SNMPv3 credentials, plus an optional trap destination for push alerts on drive failure.
Expect controller status, per-disk health, and RAID/virtual-disk state as the core metrics; a predictive drive failure warning here is exactly the kind of signal worth acting on before it becomes a rebuild.