• Some users have recently had their accounts hijacked. It seems that the now defunct EVGA forums might have compromised your password there and seems many are using the same PW here. We would suggest you UPDATE YOUR PASSWORD and TURN ON 2FA for your account here to further secure it. None of the compromised accounts had 2FA turned on.
    Once you have enabled 2FA, your account will be updated soon to show a badge, letting other members know that you use 2FA to protect your account. This should be beneficial for everyone that uses FSFT.

Management Interface And Standby Adapters (ESXi 5)

A

Adam12176

Limp Gawd
Joined
Dec 27, 2007
Messages
205
Back again! I think/hope this one is less hardware related, and more about a logical piece I'm missing.

I have two interfaces I'm trying to assign as active/standby and standby/active for Management and vMotion respectively. For a better graphic than I can provide, see Duncan Epping's article here. I stumbled across that while trying to verify what I was trying to do wasn't insane. That graphic is exactly what I was trying to do.

So - the issue is that when the second interface is added to the vSwitch and added as a standby adapter for the port group, the hosts loses connectivity to vCenter. Doesn't ping. Both interfaces are listed as active in DCUI. Honestly, I didn't think to look at this before and after. I don't know if a standby adapter in DCUI should be listed as 'used'. Verified physical wiring too, that port is absolutely cabled where I would expect it to be.

The article above does reference setting failback to 'no' to avoid a false positive at boot. I haven't tried this yet. Any ideas? Do other people have this set up a different way that I haven't considered?
 
Do you have the two adapters set in a port-channel on the physical switch? If so, don't.
 
No, those two NICs aren't teamed or anything.

I just tried this again to make sure I didn't miss something. After setting the second NIC to standby, ESX won't reboot. It sits there indefinitely saying that a reboot is in progress, but won't actually do it.

Updating some other crap (DRAC, Lifecycle) now - we'll see what the score is in a few minutes when it comes back up.
 
Adam12176 said:
Dell R810

One Broadcom 5709 interface, and one Intel ET2.
Click to expand...

wouldn't that be TWO 5709's ? and one of them is the iDrac6 port.. ?


And im sure the R810 has 4 ethernet ports ?

r810-back.jpg
 
Yes, Dash.

There are 4 integrated Broadcom interfaces, and I have one Intel ET2 quad nic in the machine too. The idea is to use one interface from each NIC to protect against failure of the card/integrated interfaces.

The problem is that when you set the Intel NIC to standby (or active, tried that too now) under network config. When I reboot the machine it won't connect to anything. If you remove the second NIC in DCUI everything goes back to normal. Even better, I can add the second NIC back in as a secondary and ESX isn't reporting that host as down. It's only on reboot, and removing/readding the adapter fixes it. This is really strange.

NetJunkie, would disabling failback be worth a shot on another reboot? I'm just wondering if that is relevant given Duncan's stance on it.
 
Yeah, understood. This is just getting really weird.

Did one more impromptu test before I left: Unplugged the Broadcom interface and sure enough, the Intel one took over. Host still pings, reported as up in vCenter. So as far as I can tell the actual interfaces are working correctly, just the ESX part of this equation isn't doing something right on startup.

I might try it with two Broadcom interfaces tomorrow, see if I get the same behavior.
 
Works with two Broadcom interfaces. No intervention required on reboot. Isn't that a beeyotch.
 
You must log in or register to reply here.
Back
Top