I was installing a new ESX Host today, and before putting it into Production, I wanted to test that the Network Failover was working correctly.
Now instead of going over to the server and pulling out network cables, I started to think, there must be a way to do this (like all good administrators do) from the comfort of my desk.
Easiest would be – Log into the switch and shutdown the port. I agree, but what happens when your the Virtualization Team do not control the Physical network (my situation)? That was not an option.
I started to look if I could bring the interface down.
With regular Linux you would – ifdown eth0
[root@esx1 ~]# ifdown eth0
usage: ifdown <device name>
So that did not work! Pretty obvious why though… There is no eth0 interface.
[root@esx1 ~]# ifconfig
lo Link encap:Local Loopback
inet addr:127.0.0.1 Mask:255.0.0.0
UP LOOPBACK RUNNING MTU:16436 Metric:1
RX packets:784233 errors:0 dropped:0 overruns:0 frame:0
TX packets:784233 errors:0 dropped:0 overruns:0 carrier:0
collisions:0 txqueuelen:0
RX bytes:3017478364 (2877.6 Mb) ...:3017478364 (2877.6 Mb)
vmnic0 Link encap:Ethernet HWaddr 00:1A:64:E5:31:38
UP BROADCAST MULTICAST MTU:1500 Metric:1
RX packets:6602129 errors:0 dropped:0 overruns:0 frame:0
TX packets:237757 errors:0 dropped:0 overruns:0 carrier:0
collisions:0 txqueuelen:1000
RX bytes:1130250120 (1077.8 Mb) ..es:195027359 (185.9 Mb)
Interrupt:113 Memory:92000000-92012100
vmnic1 Link encap:Ethernet HWaddr 00:1A:64:E5:31:3A
UP BROADCAST MULTICAST MTU:1500 Metric:1
RX packets:505 errors:0 dropped:0 overruns:0 frame:0
TX packets:0 errors:0 dropped:0 overruns:0 carrier:0
collisions:0 txqueuelen:1000
RX bytes:303685 (296.5 Kb) TX bytes:0 (0.0 b)
Interrupt:201 Memory:94000000-94012100
vswif0 Link encap:Ethernet HWaddr 00:50:56:4C:43:11
inet addr:1.1.1.1 Bcast:1.1.1.255 Mask:255.255.255.0
UP BROADCAST RUNNING MULTICAST MTU:1500 Metric:1
RX packets:2490053 errors:0 dropped:0 overruns:0 frame:0
TX packets:260404 errors:0 dropped:0 overruns:0 carrier:0
collisions:0 txqueuelen:1000
RX bytes:664363188 (633.5 Mb) TX bytes:196222902 (187.1 Mb)
So I tried..
[root@esx1 ~]# ifdown vmnic0
usage: ifdown <device name>
And that did not work either.
I did not want to shutdown the whole vSwitch – because then I would lose the connectivity (of course there is the possibility of creating the second Service Console – but this was not the case here)
The only way to do this is to manipulate the NIC’s directly. So what can you do with the physical NIC’s?
[root@esx1 ~]# esxcfg-nics
esxcfg-nics <options> [nic]
-s|--speed <speed> Set the speed of this NIC to one of 10/100/1000/10000.
Requires a NIC parameter.
-d|--duplex <duplex> Set the duplex of this NIC to one of 'full' or 'half'.
Requires a NIC parameter.
-a|--auto Set speed and duplexity automatically. Requires a NIC parameter.
-l|--list Print the list of NICs and their settings.
-r|--restore Restore the nics configured speed/duplex settings (INTERNAL ONLY)
-h|--help Display this message.
So you cannot shut down the device, but you do the next best thing – set it to 10 Half-Duplex.
That will bring the port down.
[root@esx1 ~]#esxcfg-nics -s 10 -d half vmnic2
And to bring it back up
[root@esx1 ~]# esxcfg-nics -a vmnic2
Now unless there is a good reason – your ports should always be set to Auto-Negotiate, and in this casem when the nic was reset to auto – then it came back up without any issues.
The failover worked – so my experiment was successful.
Some might call it being a lazy Bum – I call it being a Good Administrator!!
Hope you enjoyed the ride!