I have a question about ESXi storage path failover performance degradation

Hello Siti Omar,

On VMware ESXi hosts, when a Fibre Channel path fails and traffic falls back to a low‑bandwidth path, you can correct this by explicitly setting the Native Multipathing Policy (NMP) to Round Robin. The configuration is done at the datastore device level. First, identify the device with esxcli storage nmp device list and note the naa.* identifier. Then apply the policy using esxcli storage nmp device set --device naa.xxxxx --psp VMW_PSP_RR. This forces the host to distribute I/O across all available active paths instead of sticking to a single path. You can verify the change with esxcli storage nmp device list again to confirm the PSP is set to VMW_PSP_RR. For persistent configuration across reboots, ensure that the claim rules are updated if needed, but in most cases setting the PSP at the device level is sufficient. This adjustment balances throughput across multiple paths and prevents bottlenecks when a primary path fails.

If my answer is useful for you, please hit Accept the answer to support me.

Thank you,

HL.