Packet loss occurs when data packets sent across a network fail to arrive. It is expressed as the percentage of packets lost over a measurement period. Causes include congested links, faulty cables or Wi-Fi interference, overloaded routers and rate-limited or filtered traffic.
Why it matters
TCP detects missing packets and retransmits them, which slows the connection and increases delay, so even 1 to 2 percent loss can noticeably reduce throughput. Real-time traffic such as video calls and games over UDP cannot wait, so the user hears or sees the gap. Sustained loss on a path often looks like intermittent timeouts or "site down from some places".
Example
ping -c 20 example.com
# 20 packets transmitted, 19 received, 5% packet loss
To find where along the path loss starts, use mtr:
mtr -rwc 100 example.com
Loss that begins at one hop and continues to the destination indicates a real problem there. Loss at a single middle hop that disappears later is usually just that router de-prioritizing ICMP replies.
Common mistakes
- Judging by a few pings. Collect dozens to hundreds of samples.
- Treating loss at an intermediate
mtrhop as the culprit when later hops are clean. - Ignoring that firewalls and rate limits can drop ICMP while real traffic works.
- Testing only from the server. Loss may exist only on the client's path.
Related
Get notified when this term updates
Plus occasional practical tips on uptime and monitoring, nothing else. Unsubscribe anytime.
यह पेज सामान्य जानकारी है, जिसे हमारी संपादकीय टीम ने लिखा है और दस्तावेज़ीकरण बदलने पर अपडेट किया जाता है। सर्वर और प्रदाता अलग-अलग होते हैं, इसलिए बदलाव पहले सुरक्षित तरीके से जाँचें। देखें: संपादकीय नीति और अस्वीकरण.