Problem between server and Virtual Connect

[Ryan Templeton]

I have (4)685s and I'm using a Flex 10 module. In Virtual Connect I have 4 virtual nics configured at 4.5GB. Doesn't look like a problem on the VC at all. When you look at the NICs on the servers, you will see all of the NICs but 2 are running at 4.5GB and the other 2 are running at 1GB. Why would this be?

[Bart_Heungens]

For what I understood, this is not normal...
If you assign too much bandwidth (like you did), normally the NICs get a percentage of what you assigned... So if you gave 4 times 4,5 Gb, it would result that all NICs would get 2,5 Gb...

Did you try to change the numbers already?

Which firmware do you have on VC level?

[Ryan Templeton]

We just upgraded to 3.18 firmware last week and it looks like this started then.

One of the Flex-10s is reporting the proper numbers but the other isn't.

I'm going to reboot it tonight to see if that fixes it.

[Bart_Heungens]

Did you try a resetvcm already with the VCSU tool? This doesn't need downtime for the blades and can solve the issue mentioned above... I am running also 3.18 and do not have this issue (checked this morning)...

[Ryan Templeton]

Ok I ran resetvcm and but that didn't correct the problem. Once VC is reporting the proper numbers while the other is still not reporting correctly.

[Bart_Heungens]

What happens when you put the NICs to Automatic? Normally then they should get all 2,5Gb?

[Ryan Templeton]

can I do that with the blades live and in a cluster?

[Bart_Heungens]

I would not do this during business hours no... You never know...

[Ryan Templeton]

Ok thanks.