Unsolved
This post is more than 5 years old
4 Posts
0
29805
February 20th, 2009 07:00
Ethernet Controller vs. Ethernet NIC - Why Both?
I'm a newbie to networking and I have questions. I have a Dell PowerEdge T300 server equipped with a Broadcom NetXtreme BCM5722 Ethernet PCI Express NIC (dual port) and an Intel 82571EB PRO/1000PT 1GbE, PCIe-4 Gigabit Ethernet Controller (dual port). Can I connect this computer to the data port at the wall through any of the four ports? Should there be a cable connected between one port of the controller and one port of the NIC? Why is one called a Network Interface Controller and the other called an Ethernet Controller? If I only wanted two ports available could I disable one device and just use the other? I told you I was a newbie - would appreciate any enlightenment you can provide.
Larry R
2 Intern
•
1.7K Posts
0
February 20th, 2009 10:00
From what you've posted, the difference is that the Broadcom is a 10/100 connection, while the Intel is 1000 Mbps (gigabit). I'm not familiar with the PoweConnect equipment, however, so I'm moving this to the server forum so folks can answer that part of your post.
CTJoyce
20 Posts
0
February 27th, 2009 18:00
As previously stated the Intel is the NIC to use in this case. It is a full duplexing GB NIC vs the Broadcom which is only 10/100. The idea behind having the two connections per card is to be able to load balance, have redundency, or to have a bridge so if one nic fails, the other picks up the slack. Personally I have all my servers setup with redundency because uptime is more important to me then speed.
Cheers
Cameron