ESXi Host Unable to Ping Storage Array IP Following Upgrade
Article ID: 450303
Updated On:
Products
Issue/Introduction
After upgrading VMware ESXi to version 9.1, the host (hostname) loses Layer 3 network connectivity to backend storage array IP addresses over VMkernel adapter vmk1.
Executing a vmkping command from the host shell results in 100% packet loss:
Environment
- VMware ESXi 9.1.0
- Dell PowerEdge DL360 Gen10
- Network Adapters: Intel I350 Gigabit Network Connection (
vmnic0), Intel Ethernet Controller X550 - Dell Unity Storage Array
Cause
Physical network cables on the ESXi host are transposed, causing VMkernel adapter vmk1 to route traffic into the physical switch port assigned to vmk2 and attempt communication with the wrong Storage Processor (SP) on the storage array
Resolution
1. Inspect physical-to-virtual interface mappings and neighbor details from the ESXi shell using CDP or LLDP
esxcli network diag cdp check -i vmnic0
2. Review top-of-rack physical switch MAC address tables to confirm which physical switch ports are receiving frames from MAC addresses
3. Re-cable the physical network connections on the ESXi host so that `vmk1` and `vmk2` uplinks connect to their designated physical switch ports and corresponding Dell Unity Storage Processor subnets.
4. Verify Layer 3 reachability from the ESXi host to the storage array
vmkping -I vmk1 -d -s 1472 IP x.x.x.x
Additional Information
- For step-by-step guidance on testing VMkernel network connectivity and isolating physical network issues, see Testing VMkernel network connectivity with the vmkping command.
- For instructions on capturing network traffic on ESXi host interfaces using the native packet capture tool, see Using the pktcap-uw tool in ESXi.
- For procedures on escalating complex support cases or raising case concerns within the Broadcom Support Portal, see Escalating Broadcom Support Cases
Feedback
