The issue is introduced in ESXi 7.0U3C and resolved in ESXi 7.0 U3L(Build- 21424296) and later releases.
Workaround: Once the ESXi Host reaches the HUNG state, you will need to forcefully reboot the ESXi Host through ESXi DCUI from the Hardware Remote Console.
Then to avoid this issue from re-occurring, please enable per file I/O scheduling following the below steps:
Connect to the ESXi Host Shell through SSH
Run the below commands
# esxcli system settings kernel set -s isPerFileSchedModelActive -v TRUE # reboot