The vmkernel alert shown indicates that ESXi has received an interrupt on an unexpected processor interrupt vector and has ignored the interrupt because ESXi does not know the source of the interrupt or which software handler to invoke.
Reception of such an interrupt is not normal and may cause loss of I/O connectivity and/or ESXi PSODs.
In order to distribute the interrupt load evenly across processors, ESXi performs interrupt re-balancing. As part of this process, ESXi may modify the target processor and interrupt vector assigned to an I/O device. This process is known as I/O device interrupt migration.
When the VT-d interrupt remapper is not present in a platform or its use is disabled in ESXi (boot option iovDisableIR=TRUE), ESXi performs I/O device interrupt migration by modifying the PCI MSI/MSI-X registers in the I/O device(s).
VMware has recently received reports that some I/O devices do not react well when ESXi performs interrupt migration by programming the device's MSI/MSI-X registers. Specifically, some I/O devices may interrupt in a processor vector which ESXi no longer uses and considers invalid, causing the generation of the alert message and its subsequent effects.