ESXi hosts lost storage paths unexpectedly for a short time
Article ID: 418143
Updated On:
Products
Issue/Introduction
Hosts lost paths to storage causing VMs to lose disk connection for a few minutes.
The issue resolved itself after some time.
Using Cisco UCS VIC Fnic Controller with firmware 4.5(3e) on a host running ESXi 8.0 U3
Environment
ESXi 8.0 U3
Cisco UCS hosts
Cause
Firmware installed on the Fnic Controller is below the minimum required version.
Minimum required firmware version is 4.6(5) while installed is 4.5(3e).
Aborts seen in logs:
2025-11-03T18:02:42.413Z In(182) vmkernel: cpu76:2097228)nfnic: <1>: INFO: fnic_fcpio_icmnd_cmpl_handler: 1865: io_req: 0x45b983403210 sc: 0x45b96bc8fb40 tag: 0x43b CMD_FLAGS: 0x53 CMD_STATE: FNIC_IOREQ_ABTS_PENDING ABTS pending hdr status: FCPIO_ABORTED scsi_status: 0x0$2025-11-03T18:02:42.382Z Wa(180) vmkwarning: cpu46:3059863)WARNING: nfnic: <1>: fnic_abort_cmd: 3866: Abort for cmd tag: 0x342 in pending state
Unexpected behavior, including path failures and commands aborting, can occur when the driver and firmware do not meet the compatibility list requirements for a given device.
Resolution
Update the Fnic controller firmware to the minimum required version as per the HCL.
Work with the hardware vendor to obtain any required firmware versions.
Additional Information
Feedback
