The VMware Knowledge Base provides support solutions, error messages and troubleshooting guides
Search the VMware Knowledge Base (KB)
View by Article ID
Change to ICMP ping response behavior in ESXi 5.1 (2042189)
Symptoms
In ESXi 5.1, it is no longer possible to ping VMkernel interfaces other than the management VMkernel interface from client machines in other networks and VLANs.
Cause
This behavior is a consequence of changes introduced in ESXi 5.1. ICMP Echo replies are now only sent back out the same interface that the Echo Request was received on.
This behavior was introduced because it is desired by multi-pathing capable applications. For example, an iSCSI or NFS storage filer may use ICMP pings to determine whether a given VMkernel is reachable or not. For this functionality to work as designed, the host must respond to these echo requests on the same VMkernel interface that they are received on.
There is no change in behavior when pinging from a source within the same subnet as the VMkernel interface, only when the ICMP Echo Reply must be routed to get back to the source.
In previous versions of ESXi, pings originating from a different subnet would simply be replied to based on the host's routing table, normally through the default gateway defined for the management interface, vmk0.
Note: This behavior is applicable only to ICMPv4 traffic. All other forms of TCP/IP traffic are routed normally via the host's routing table.
Resolution
When performing network troubleshooting with ESXi 5.1, ping VMkernel interfaces from a source machine in the same subnet.
For example, ping an ESXi 5.1 host's iSCSI VMkernel interface from the SAN itself, or from another ESXi 5.1 host with a VMkernel interface in the same subnet/VLAN.
This Article Replaces
2041264
Request a Product Feature
To request a new product feature or to provide feedback on a VMware product, please visit the Request a Product Feature page.
