Hi;
I have 5 ESXi's in a cluster, recently implementing a best practice I separate to another switch the Vkernel Port to use VMotion, 2 of the ESXi's have no problem in making a Vmotion, but in 3 of them there's an error of "cannot comunicate with the Vkernel Port" I log in to the SSH console and make the vmkping tests and there it's no comunication between them in the vkernel port, the strage thing is that not even in the same ESXi's I can't ping to his vkernel port, like if the network adapter it's down, but in the inventory of the ESXi does not look if there is a problem, the network adapter it's active.
The sad part it's that I can't restart the ESXi because there are VM's running, and there is no Vmotion to put them in another host, hope there's a solution to my problem without a ESXi restart, if there's not i have to make a maintenance window
hope you can help me..
Note: Discussion successfully moved from VMware vCenter™ to vMotion & Resource Management
Can you provide a screenshot or diagram of how the networking is or was changed?
Before
Switch 3
VkernelPort1(conection to SAN)
VKernelPort2(conection to SAN)
VKernelPort3(conection to SAN)
VKernelPort4(conection to SAN)(Vmotion enable)
AFTER
Switch 3
VkernelPort1(conection to SAN)
VKernelPort2(conection to SAN)
VKernelPort3(conection to SAN)
Switch 4
VKernelPort4(Vmotion enable)
What about the physical networking that is used with the vSwitches?
check ur vlan tagging and also make sure in the vswitch for the vmotion u do see the observed ip range properly
Hi, thanks for your response;
There is no VLAN Tag because it's the native VLAN, so I left it in "0(none)" I even tag it to the VLAN of the network (that to me is VLAN:250) and nothing.
It comes to my attention that you said that the Vswitch for the Vmotion sees the iprange, the thin is that i DON't see the iprange, to remind my escenario:
Before
Switch2
vmkernelnic9 vmnic9 1000 full
vmkernelnic8 vmnic8 1000 full
vmkernelnic7 vmnic7 1000 full (vmotion)
vmkernelnic6 vmnic6 1000 full
After
Switch2
vmkernelnic9 vmnic9 1000 full
vmkernelnic8 vmnic8 1000 full
vmkernelnic7 vmnic7 1000 full
Switch3
vmkernelnic6 vmnic6 1000 full (vmotion)
At first it work fine, but i recently had problems migrating a VM, then i noticed that the vmkernel of the vmkernelni6 stop working at a migration, then i did a vmkping test and there is no response
The before scenario of the Swith2 was for a iSCSI software adpter using por binding, then the "After" scenario it's by a best practice of VMware, separating the traffic of the SAN and the Vmotions.
After the failed vkernel (switch 3), i just enable the vmotion in the vmkernelnic7 and works fine, but in neither the vSwith2 or vSwithc3 there are observerd IP ranges present:
Hi;
I reboot the Host...NOTHING
Just added another nic from the switch2 and it started pining, now I have 2 network adapters for the conection to the SAN and 2 for the Vmotion Switch, and now it's working, but my goal it's to make it work with only 1 network adapter for the vmotion so the Switch2 has 3 network adapters for the traffic of the SAN connection...