What is Packet Loss?
Learn what packet loss is, what causes it, and how TCP vs UDP applications handle it — with networking interview questions.
Expected Interview Answer
Packet loss is the failure of one or more transmitted packets to reach their destination, caused by congestion, faulty hardware, signal interference, or a router intentionally dropping packets when a queue overflows.
Every network device has finite buffer space; when packets arrive faster than a link can forward them, the queue fills up and the router discards the excess, a mechanism called tail drop. Wireless interference, damaged cabling, overloaded switches, and misconfigured QoS policies are other common causes. TCP detects loss through missing acknowledgements or duplicate ACKs and retransmits the lost segment, trading latency for reliability, while UDP-based applications like video calls simply tolerate or conceal the gap. Persistent packet loss above roughly one to two percent noticeably degrades throughput and interactive application quality, so operators monitor loss rates as a core network health metric.
- Signals congestion or faulty links before users notice
- TCP retransmission masks loss at the cost of latency
- Loss rate is a key SLA and monitoring metric
- Diagnosing loss location narrows down root cause
AI Mentor Explanation
Packet loss is like a fielder relay throw that never reaches the wicketkeeper because the outfielder overthrows into the crowd — the run-out attempt fails and the team must react, either accepting the missed opportunity or trying again next ball. When too many relay throws go astray in an innings, the fielding side’s overall effectiveness collapses. Cricket captains watch the fielding error rate the same way network engineers watch packet loss percentage.
Step-by-Step Explanation
Step 1
Congestion builds
A link or device receives more traffic than it can forward, filling its output queue.
Step 2
Buffer overflows
Once the queue is full, the device drops (tail-drops) additional incoming packets.
Step 3
Detection
TCP notices missing or duplicate ACKs; UDP applications notice gaps in the expected stream.
Step 4
Recovery or degradation
TCP retransmits the lost segment; loss-tolerant apps conceal or skip the gap instead.
What Interviewer Expects
- Explains congestion/buffer overflow as the primary cause
- Distinguishes how TCP vs UDP applications react to loss
- Names other causes: interference, faulty hardware, misconfiguration
- Knows loss rate is a standard monitoring/SLA metric
Common Mistakes
- Assuming packet loss always means a broken link
- Not knowing TCP automatically retransmits lost segments
- Confusing packet loss with latency or jitter
- Thinking UDP applications always fail outright on loss
Best Answer (HR Friendly)
“Packet loss is when some of the data sent over a network never arrives, usually because a router got overloaded and had to drop it. Reliable protocols like TCP notice and resend the missing piece, while things like video calls just glitch briefly instead. A little loss is normal, but consistently high loss makes everything feel slow or broken.”
Code Example
# Send 50 pings and summarize loss percentage
ping -c 50 8.8.8.8 | tail -3
# 50 packets transmitted, 47 received, 6% packet loss
# Trace the path and see per-hop loss (requires mtr)
mtr --report --report-cycles 50 example.comFollow-up Questions
- How does TCP detect and recover from packet loss?
- What is the difference between packet loss, latency, and jitter?
- How does congestion control respond to packet loss?
- Why do real-time applications tolerate packet loss instead of retransmitting?
MCQ Practice
1. What is the most common cause of packet loss on a busy network?
When a device receives more traffic than it can forward, its output buffer fills and excess packets are dropped.
2. How does TCP typically respond to detected packet loss?
TCP treats missing or duplicate acknowledgements as a loss signal and retransmits the affected segment.
3. Which metric is most directly used to monitor packet loss?
Packet loss is expressed as the percentage of transmitted packets that never arrive at the destination.
Flash Cards
What is packet loss? — Packets that fail to reach their destination, often due to congestion-driven buffer drops.
How does TCP handle loss? — It detects missing ACKs and retransmits the lost segment.
How do UDP apps handle loss? — They tolerate or conceal gaps instead of retransmitting.
What tool measures loss per hop? — mtr, which combines traceroute and ping statistics.