1. Check the switch configuration.
The port configurations of the switch connecting to the faulty CNA node are the same as those for connecting to normal CNA nodes. Therefore, the switch fault is ruled out.
2. Check the network interface card (NIC) connecting to the storage.
NICs of normal CNA nodes are 1000 MB NICs while the NIC of the faulty CNA node is a 100 MB NIC.
Run the ethtool command to check the speed of the NIC, it's showing as bellow.
3. Replace the NIC of the faulty CNA node with a 1000 MB NIC and test the performance of VMs on the CNA node. All VMs function properly with normal performance.
The NIC cards connecting to the storage array are 100MB.