BladeSystem - General
Showing results for 
Search instead for 
Did you mean: 

strange NIC teaming behaviour

Honored Contributor

strange NIC teaming behaviour

Our environment: BL460 G1/G6 servers, C7000 enclosures, VC modules as interconnect.
OS: Windows 2008 R2

All of the latest Firmware/Drivers/PSP are installed.
We are using the two onboard NICs in a Team configured for "Transmit Load Balancing with Fault Tolerance (TLB)"

NetApp 3140 cluster: Controller #1 - IP Address #1, Controller #2 – IP Address #2
Both IP's in same subnet.

Problem: When we try to connect via telnet to IP Address #1 - we can establish a connection immediately, but when we try to telnet to IP Address #2, we only get a reply/connection after about 5 - 15 seconds. Then, if we try to telnet to IP Address #2 first, we connect immediately, but after that if we try to connect via telnet to IP Address #1 we again need to wait 5 – 15 seconds in order to be able to connect. The same thing happens with SSH using Putty when we try. We are using the built-in Windows telnet client, no third-party software.

If we break the team, we get an answer immediately from both IP's, regardless of which order we try to connect.

We experience the same problem with other network devices, although it is mostly relevant to us with our NetApp – but you can replace it with any other network device.

We also see this issue when the Team Type is set to use "Network Fault Tolerance Only (NFT)". We tried numerous time to play around with the various Transmit Load Balancing methods but they all basically didn't help us resolve this particular issue.

Actually, we have same problem when we create teaming on DL360 G5 as well, so problem is not related to Virtual connect.

Any suggestions ?

Thanks, Igor