• 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.

ESXi link aggregation

D

Dami

Weaksauce
Joined
Mar 28, 2010
Messages
97
I have a problem setting link aggregation on my esxi server. Its a Supermicro X9SRL-F with dual nics connected to a Cisco SG300-20 switch. I configured the host as described here and when I put the 2 switch ports in LAG group (IP/MAC Load Balance Algorithm, and no LACP as I know that this requires vDS and I dont really need it), the management host is no longer reacheble, but other VMs work. I have a few VLANs on the switch but not on the host or my desktop ports (vlan1 is the default on all the ports anyway). I did put both groups to use all of the vlans (0-4096) but that didn't help.
Am I missing something else?
 
Did you set the load balancing policy to Route Based on IP Hash on the vSwitch side?
 
Ofc i did as this is basically the only thing i knew to do on the host side.
 
Can you supply a screenshot of how you have it setup on the switch and Host side?
 
If you set your port-channel to use IP based hashing you need to set the same on your vSwitch. For MAC based hashing you do nothing on either end. You put all the active NICs you want in the vSwitch, set them to Active, and the switch doesn't even need to know. But for IP hashing... The switch has to have those ports in a port-channel, be set to IP hashing (which should be the default in a port-channel on that switch), and you have to set the vSwitch to use IP hashing.

I don't do that in my lab..mainly because I don't need IP hashing. I don't have a bunch of clients talking to a server VM...and it's more hassle to add/remove NICs from the vSwitch since you have to make changes on the vSphere side as well as the switch side. I use MAC-based hashing or when I'm using the VDS I use load-based teaming..which underneath is really just Virtual Port ID..which is almost the same as MAC-based hashing.
 
Make sure the management network if configured to use Link Aggr as well. You have to set it for both if the management network is using the same network as the VMs. And it needs to be set to IP hashing as well
 
m1abram said:
Make sure the management network if configured to use Link Aggr as well. You have to set it for both if the management network is using the same network as the VMs. And it needs to be set to IP hashing as well
Click to expand...

I cant test it just yet, but I think this was the problem. I only set the vswitch to IP hashing, which also set the VM network group but not the management. Thanks everyone for the help and geiger for the great guide link.
 
A quick way to see if your hashing is wrong. Try pinging the IP of the VMkernel from a few different hosts. I bet a couple work, and the others don't. Sometimes with incorrect hashing you get lucky and it lines up anyway. So if you ever strand a host like this don't do anything drastic right away. First try to connect from a few systems and see if one can connect using the VI client to fix it.
 
You must log in or register to reply here.
Back
Top