A server in a data center is running a critical application. The server is connected to a managed switch via two network cables for redundancy (link aggregation). Recently, the server experienced intermittent connectivity issues. The network logs show that every few hours, one of the links goes down and then comes back up within a minute. The other link remains up. The server's team reports no errors in the server logs. What is the most likely cause of this issue?
Poor termination can cause intermittent connectivity on a single cable.
Why this answer
Option C is correct because a poorly terminated cable can cause intermittent physical-layer issues such as impedance mismatches or partial breaks, which lead to link flaps (link down/up events) without generating errors in the server's operating system logs. In a link aggregation group (LAG) using LACP or static aggregation, a single faulty cable will cause only that link to drop, while the other link continues to carry traffic, matching the described symptom of one link cycling every few hours.
Exam trap
The trap here is that candidates often assume software or configuration issues (like broadcast storms or adapter overheating) when the symptom of a single link flapping in a LAG points directly to a physical-layer problem with that specific cable, not a systemic or hardware-wide fault.
How to eliminate wrong answers
Option A is wrong because a faulty switch uplink port would affect all traffic passing through that uplink, not just one specific link in the aggregation group, and would likely cause connectivity loss for multiple devices or the entire switch. Option B is wrong because a broadcast storm would overwhelm the network with broadcast frames, causing both links to be saturated and the server to experience general connectivity loss, not just one link flapping while the other remains stable. Option D is wrong because an overheating network adapter typically causes the entire adapter to fail or throttle, affecting both links simultaneously (since they share the same physical adapter), and would likely produce temperature-related errors in the server logs.