It is worth noting that network interface cards (NICs or also ethernet cards) are one of the rarest failures. In the vast majority of times a server goes offline, it is more likely to be an issue higher up on the network (blackholing, null routing, bad cable, etc.) than a failed NIC. However there are some symptoms that can help determine a failing or bad NIC. Also note that you will in most cases need to request support with an issue like this since a faulty NIC will prevent the server from coming online.
If a server goes offline, the first thing to do is submit a support ticket via MyCP notifying us of the problem. We will go down a list of common causes for a server going offline and determine whether the NIC is bad. Before we reach this conclusion however, we will need to check the physical cabling, the status of the link/activity lights on the NIC, the switchport configuration, and the link light on our network access switch. The second most likely cause for a server being offline is the network configuration within the operating system which is verified against the proper network configuration.
If everything else checks out on the server and higher up in the network we will then replace the NIC card (if it is in a PCI expansion slot), install a PCI expansion card (if the motherboard port is bad), or in extreme cases where other components on the motherboard are showing signs of failure we will replace the entire motherboard.
If you suspect faulty network hardware on your server you should immediately open a support ticket via MyCP. We have many tools at our disposal to narrow down and repair the issue and are more than happy to assist free of charge (on rented SLS servers) with hardware replacement.