VMs in the same subnet hosted on different ESXi hosts are unable to communicate with each other
Article ID: 417824
Updated On:
Products
VMware vCenter Server
Issue/Introduction
- VMs in the same subnet and VLAN hosted on different ESXi hosts are unable to ping each other.
- Initiate the pings from source VM to destination VM and the pings do not go through:
Environment
VMware vCenter Server
VMware ESXi
Cause
- In the packet captures performed on uplink used by source and destination VM, its observed that while pinging the destination VM from source VM, the ARP requests are leaving the source vm's host uplink but no ARP requests are coming on the uplink of destination VM's host.
- Below command can be used to perform packet captures on uplink of ESXi host used by source and destination VMs:
pktcap-uw --uplink vmnicx --capture UplinkSndKernel,UplinkRcvKernel -o - | tcpdump-uw -r - -enn- pktcap-uw --switchport <switchportid> --capture VnicTx,VnicRx -o -| tcpdump-uw -r - -enn
- Replace vmnicX with the appropriate uplink interface used by VM .
- For detailed steps on performing packet captures, refer to the Broadcom KB article:https://knowledge.broadcom.com/external/article/341568/using-the-pktcapuw-tool-in-esxi-55-and-l.html
Resolution
Since the packet drop is identified as occurring outside the ESXi hosts on the physical network, the issue must be resolved by the physical network team.
-
Engage the physical network team.
-
Request the team to check the physical network configuration (e.g., VLAN settings, switch port configuration, spanning tree) relevant to the two ESXi hosts' uplinks.
-
Instruct the team to trace the packets in the physical network infrastructure to precisely identify the drop point and correct the misconfiguration.
Additional Information
Feedback
Yes
No
Powered by
