In IP networking, what is a “fragment” and in what context does it occur?

Introduction / Context:
IP fragmentation allows oversized datagrams to cross links with smaller Maximum Transmission Units (MTUs). Understanding what a fragment is and why fragmentation happens is essential for troubleshooting path MTU issues and reassembly failures.

Given Data / Assumptions:

• An IP datagram exceeds the MTU of some link on its path.
• An IP router or the source host may split the datagram into pieces.
• Fragments are reassembled by the destination host using identification and offset fields.

Concept / Approach:

A fragment is not a protocol mechanism to stop looping (that is handled by fields such as TTL) and is not encapsulation (placing a protocol data unit inside another). It is specifically a smaller slice of the original datagram produced to satisfy MTU limits.

Step-by-Step Solution:

Verification / Alternative check:

Packet captures show multiple fragments with identical identification values and increasing offsets when fragmentation occurs, confirming the definition.

Why Other Options Are Wrong:

Best-effort looping avoidance: That refers to TTL or hop limit, not fragments.

Encapsulation: Describes layering, not breaking a datagram into pieces.

All of the above / None of the above: Incorrect because the specific, correct definition is option C.

Common Pitfalls:

Confusing fragmentation with segmentation (transport-layer) or with tunneling/encapsulation.

Final Answer:

One of the smaller pieces created when a gateway divides a large IP datagram to traverse a network with a smaller MTU