Ttdfjt

I have two servers, the first with a built-in and add-on 10gbit port, and the second with built-in gbit and add-on 10gbit (Syba AQC107), communicating through a feed-through rated for gigabit communication (Cat6A shielded cabling totaling ~50ft, unshielded couplers). Thus, I don't expect full 10gbit communication to work. However, the two built-in ports communicate reliably and I get full gigabit transfer speeds. When one of the Syba ports is connected at either end, the autonegotiation is unreliable, and usually falls back to 100Mbit. However, when it does succeed at gbit autonegotiation, it does get full gigabit transfer speeds. When both Syba cards are plugged in, no link is detected at all. Connecting the Syba cards directly (50ft cable, no feedthrough) results in 10gbit throughput.

I infer from this that the current constraint is the Syba port's autonegotiation capability, and not the physical cabling. Thus, I hope that I may be able to at least get 2.5Gbit transfer speeds with a different NIC, but I don't have any way to know a priori that a given card will be suitable without the impedance, line filtering, etc. of the port. Is this simply an instance of money=power? Does anyone have brand suggestions that might prove fruitful?

Related, how is it possible that these cards can reliably get gigabit throughput, but frequently (~80% of time) fail to autonegotiate gigabit speeds? Is autonegotiation typically handled by a different DSP?

I am using ethtool for NIC status and iperf for transfer rates.

You simply cannot mix shielded and unshielded parts in a link. The shield only works if it continuous end-to-end and properly grounded on both ends. The actual wires in shielded cabling cannot meet specifications without working shielding. Having unshielded connectors means that the shielding is broken on the link, and it is not properly grounded. Improperly grounded links with shielded cable cannot meet the required cable specifications.

There are documents that explain things for you. For example, Shielded and unshielded twisted-pair cable revisited:

If STP cable is combined with improperly shielded connectors,
connecting hardware or outlets, or if the foil shield itself is
damaged, overall signal quality will be degraded. This, in turn, can
result in degraded emission and immunity performance. Therefore, for a
shielded cabling system to totally reduce interference, every
component within that system must be fully and seamlessly shielded, as
well as properly installed and maintained.

An STP cabling system also requires good grounding and earthing
practices because of the presence of the shield. An improperly
grounded system can be a primary source of emissions and interference.
Whether this ground is at one end or both ends of the cable run
depends on the frequency at which a given application is running. For
high-frequency signals, an STP cabling system must be grounded, at
minimum, at both ends of the cable run, and it must be continuous. A
shield grounded at only one end is not effective against
magnetic-field interference.

It seems that the installation was poorly and incorrectly done. For the devices to negotiate at 1000Base-T (1 Gbps), all four pairs must be working correctly. Intermittent negotiation to 100Base-TX (100 Mbps) tells me that at least one of the pairs has an intermittent connection.

Your cable installer should have run the Category-6a test suite* and provided a report for each cable run. That seems unlikely because he did not correctly install the cabling (mixing shielded and unshielded parts, which cannot meet the Category-6a test suite, and probably would not even pass the Category-5e test suite). You should demand that he return to fix the installation, and if he is a certified installer, you should report the problems to BICSI.

Category-6 and Category-6a cabling is something with which even experienced installers have problems. It is critical that the finished installation be tested with a proper (expensive, several thousand dollar) cable tester, and that you receive reports for each cable run.

*This answer explains the primary test suite.