IPv4 Fragmentation Numerical Example 2

You’re working with a datagram of 3000 bytes. Out of this, 20 bytes is the IP header. So the actual payload is 2980 bytes.Now the router has to forward this packet to a link where the MTU is 500 bytes. That 500 includes the header. So each fragment can carry only 480 bytes of data because 500 minus 20 leaves 480 bytes for payload.

Understanding the given data

Step by step fragmentation

Now divide the total payload by how much each fragment can carry.

2980 divided by 480 gives 6 full fragments and some remaining data.

480 multiplied by 6 gives 2880 bytes
Remaining data becomes 2980 minus 2880 which is 100 bytes

So total fragments will be 7
First 6 fragments carry 480 bytes each
Last fragment carries 100 bytes

Each fragment again gets its own 20 byte header added

Alternative percentage method

You can also think of it this way

2980 divided by 480 is about 6.2
That means 6 full fragments and about 20 percent remaining

20 percent of 480 is close to 100 bytes
So the last fragment carries 100 bytes

Different way of thinking. Same result.

Offset calculation

Offset tells where each fragment starts in the original data. It is measured in units of 8 bytes.

So divide starting byte by 8

Fragment 1 starts at 0 → offset 0
Fragment 2 starts at 480 → 480 ÷ 8 = 60
Fragment 3 starts at 960 → 960 ÷ 8 = 120
Fragment 4 starts at 1440 → 1440 ÷ 8 = 180
Fragment 5 starts at 1920 → 1920 ÷ 8 = 240
Fragment 6 starts at 2400 → 2400 ÷ 8 = 300
Fragment 7 starts at 2880 → 2880 ÷ 8 = 360

Final result in table form

Fragment Data Bytes Identification (ID) Offset (Bytes) Offset (/8) MF Flag Total Length
1st Fragment 480 777 0 0 1 500
2nd Fragment 480 777 480 60 1 500
3rd Fragment 480 777 960 120 1 500
4th Fragment 480 777 1440 180 1 500
5th Fragment 480 777 1920 240 1 500
6th Fragment 480 777 2400 300 1 500
7th Fragment 100 777 2880 360 0 120

Final takeaway

Once MTU gets smaller, fragmentation increases quickly. More fragments means more headers, more processing, and higher chance of packet loss. So real networks try to avoid this situation. But for exams and concepts, this is exactly how fragmentation works.