Packet Fragmentation and Reassembly
See how an oversized IPv4 datagram is split, carried, and rebuilt
Packet Fragmentation and Reassembly
Follow an oversized IPv4 datagram as it is split into MTU-sized fragments. Inspect offset units, More Fragments flags, arrival order, and the reassembly result.
Fragmentation Workbench
A packet larger than the path MTU is split into smaller IP fragments.
Each link can carry only a maximum frame payload. MTU limits the IP datagram size.
Step through normal delivery, then switch to a missing-fragment failure.
Fragment offsets are in 8-byte units, not raw byte numbers.
Original datagram
The packet is larger than the MTU, so the IPv4 payload must be split into fragment payloads.
Datagram and fragments
Selected fragment fields
Arrival order
Reassembly slots
Learning Support
Learning Support
Start with MTU
If the original datagram is less than or equal to the MTU, IP fragmentation is not needed.
Subtract header
Each IPv4 fragment has its own IP header, so payload capacity is MTU minus header bytes.
Use 8-byte units
The IPv4 fragment offset field counts blocks of 8 payload bytes. Non-final fragments must align to that size.
Missing hurts
The receiver can reassemble out-of-order fragments, but one missing piece makes the whole datagram unusable.
Quick Reference
Quick Reference
Fragment payload
floor((MTU - IP header) / 8) * 8Maximum payload bytes for each non-final IPv4 fragment.
Offset field
offset bytes / 8Fragment offset points to the payload byte position in 8-byte units.
MF flag
1 except final fragmentMore Fragments tells the receiver whether another fragment should follow.
Identification
same ID on all fragmentsThe receiver groups fragments with the same source, destination, protocol, and identification.
IPv6 note
routers do not fragment IPv6IPv6 fragmentation is performed by the source, not by routers along the path.
IoT practice
avoid fragmentation when possibleSmall wireless links lose more when one fragment failure discards the whole packet.