Define the failing path
Record the host HBA, switch port, fabric, target port, VSAN, expected WWPNs, and affected LUNs. Determine whether the problem is one path, one fabric, one host, or multiple devices.
Check physical and interface state
Review interface status, speed negotiation, counters, transceiver information, link resets, and errors. A link that is up can still have physical-layer problems, so compare counters over time rather than relying only on current state.
Verify fabric login and name-server visibility
Confirm the initiator and target WWPNs are logged into the expected fabric and VSAN. If a device is missing, work backward through HBA, cable, optics, switch port, and target connectivity before focusing on zoning.
Validate zoning carefully
Confirm the active zoneset contains the intended initiator-to-target membership and that changes were activated in the correct VSAN. Prefer controlled, documented zoning changes and avoid broad zones that expose unnecessary targets.
Finish at the host multipathing layer
Once the fabric is healthy, confirm the host sees the expected target paths and that multipathing software reports the correct path count and state. Storage presentation, host rescans, and device-mapper or native multipathing can still be the final missing piece.
This article is a general troubleshooting framework. Validate commands and procedures against your platform version, vendor documentation, support requirements, and change-control process before making production changes.