What is Packet Loss?
What is packet loss? Understanding this phrase starts with packets, which are the fundamental units of data transmission necessary within computer networking. They carry information across the Internet and other networks, allowing data to be efficiently routed from one point to another.
Each data packet contains a specific amount of data, along with control information necessary for successful delivery. However, sometimes these packets don’t reach their intended destination. This phenomenon is known as packet loss.
Packet loss is closely associated with quality of service considerations. The amount of packet loss that is acceptable depends on the type of data being sent. Low levels of packet loss (less than 1-2%) are considered acceptable for most applications, but higher levels of packet loss can cause significant issues with network performance and user experience. For example, real-time applications like voice and video will be affected more severely by packet loss. Something as small as a 2% packet loss is usually quite noticeable to a listener or viewer.
Common Causes of Packet Loss
The primary cause of network packet loss is congestion. All networks have space limitations, so in simple terms, network congestion is very much the same as peak hour traffic. However, packet loss can be caused by a range of other factors such as:
- Network hardware problems
- Faulty connections
- Software errors/bugs
- High network traffic levels
- Inadequate bandwidth
- Network misconfigurations
- Overtaxed devices
Problems Caused by Packet Loss
The impact of packet loss can be significant, particularly in real-time applications and services. Voice and video communications, online gaming, and streaming media are highly sensitive to packet loss. When packets are lost, the information they carry is either delayed or permanently lost, resulting in distorted audio and video, choppy connections, or interrupted streams. In some cases, packet loss can even lead to dropped calls, freezing screens, or completely failed data transfers.
These issues negatively reduce employee productivity and satisfaction and can also affect customer experience. Ultimately, this leads to increased costs due to lowered employee productivity and often a need to invest in additional IT expenses in an attempt to mitigate these issues.
Monitoring Packet Loss
Diagnosing packet loss typically involves conducting network tests and analyzing the results. Here are several examples of ways to diagnose packet loss:
- Ping Test: A ping test can check for packet loss between a device and its destination server or website. If you see “Request timed out” or “Destination host unreachable” messages, it indicates packet loss.
- Traceroute: The traceroute tool can identify the network path and pinpoint the location where packet loss might be occurring. The traceroute will display a list of routers or hops between the device and the destination, along with the response times. Look for routers with high response times or asterisks (*) indicating packet loss.
- Network Monitoring Tools: Utilize network monitoring tools to get more detailed information about packet loss. There are various network monitoring solutions that can monitor network performance and provide insights into packet loss, latency, and other network metrics.
Measuring Packet Loss
Packet loss is typically expressed as a percentage, indicating the ratio of lost packets to the total number of packets sent. Measuring packet loss involves analyzing the network traffic and observing the absence of expected packets. This can be done using specialized tools that track packet transmission and reception, providing insights into the overall network performance.
Some methods and metrics used to quantify packet loss are as follows:
- Percentage-based measurement: Packet loss is often expressed as a percentage, indicating the ratio of lost packets to the total number of packets sent.
- Round-Trip Time (RTT): Another way to measure packet loss is by analyzing the Round-Trip Time, which is the time taken for a packet to travel from the source to the destination and back. By comparing the time taken for packets to be sent and received, any delays or missing packets can be identified as potential packet loss.
- Internet Control Message Protocol (ICMP): ICMP is a protocol used for diagnostic purposes in IP networks. It includes tools like ping and traceroute that can be used to measure packet loss.
- Jitter measurement: Jitter refers to the variation in packet delay or latency. Excessive network jitter can lead to packet loss. Tools and algorithms that measure jitter can indirectly detect packet loss by analyzing the inconsistencies in packet arrival times.
- Packet Capture and Analysis: Network administrators can use packet capture tools to capture network traffic and examine the packet headers. By inspecting these captured packets, they can identify any missing or incomplete packets, thereby detecting packet loss.
It’s important to note that different measurement methods have their own strengths and limitations. The choice of measurement technique depends on the specific requirements and the nature of the network being monitored. Network administrators often employ a combination of these methods to gain a comprehensive understanding of packet loss within their network.
How Different Protocols Affect Packet Loss
Transmission Control Protocol (TCP) and User Datagram Protocol (UDP) are two widely used transport layer protocols in computer networks. TCP provides reliable and ordered delivery of packets by establishing a connection and implementing acknowledgment mechanisms. On the other hand, UDP is a connectionless protocol that offers a lightweight and fast approach for transmitting packets without error-checking or retransmission and is commonly used for time-sensitive applications like gaming and playing videos.
TCP and UDP handle packet loss differently due to their inherent characteristics. TCP assumes that packet loss is due to congestion and applies congestion control algorithms to minimize loss. It employs retransmission mechanisms to ensure reliable delivery, which introduces some latency. UDP, however, does not provide mechanisms for retransmission or congestion control. Therefore, if packets are lost during UDP transmission, they are not automatically recovered, potentially resulting in missing or corrupted data.
Preventing Packet Loss
While it may not be possible to eliminate packet loss entirely, several measures can help mitigate its effects. Here are some suggestions:
- Optimize network infrastructure: Make sure all your infrastructure – including routers, switches, and cables – is in good condition as faulty hardware can be a contributing factor to packet loss.
- Reduce network congestion: To alleviate packet loss you can use the following techniques:
- Quality of Service (QoS) – Deploy QoS mechanisms to prioritize traffic and allocate resources. This makes sure that critical traffic like voice and video receive high priority.
- Traffic Shaping – Implement traffic shaping policies to control the rate at which packets are transmitted.
- Optimize network settings and configurations: Adjusting certain network settings can help. Here are a couple of examples:
- Adjust the Maximum Transmission Unit (MTU). The MTU is the maximum size of a packet that can be transmitted over a network. Reducing the size can prevent fragmentation and decrease the likelihood of packet loss.
- Tweak TCP/IP settings. Modifying parameters such as the TCP window size of the retransmission timeout (RTO) can improve network performance and reduce packet loss.
- Implement error detection and correction mechanisms: Use error detection and correction techniques.
- Forward Error Correction (FEC) – This adds redundant information to packets, allowing the receiver to detect and correct errors eliminating the need for retransmission.
- Automatic Repeat reQuest (ARQ) – ARQ protocols like Selective Repeat or Go-Back-N enable the receiver to request retransmission of lost packets.
- Utilize reliable protocols and technologies: Certain protocols and technologies are designed to minimize packet loss:
- Reliable transport protocols – Consider using more reliable protocols like TCP instead of unreliable protocols like UDP when data integrity and delivery are crucial.
- Packet loss concealment – In real-time applications such as voice or video communication, implementing packet loss concealment techniques can help mitigate the impact of lost packets. These techniques involve filling in missing data or employing interpolation algorithms to minimize the perceived impact of packet loss.
- Monitor and analyze network performance: Regularly monitor your network for performance issues and packet loss. Use network monitoring tools to track network metrics and pinpoint the causes of packet loss.
Using Network Management Tools to Detect & Prevent Packet Loss
Network performance management (NPM) systems can play a crucial role in detecting and preventing packet loss by actively monitoring network traffic, analyzing performance metrics, and implementing corrective measures. Here are some ways in which network performance management systems can help:
- Packet Loss Monitoring: Network performance management systems continuously monitor the network for packet loss by analyzing network traffic. They track the number of packets sent and received and compare them to identify missing packets. which indicates packet loss.
- Real-Time Performance Analysis: These systems collect and analyze various performance metrics such as latency, jitter, and round-trip time (RTT). Patterns and anomalies might indicate packet loss.
- Root Cause Analysis: NPM systems can perform root cause analysis to identify underlying reasons.
- Quality of Service (QoS) Monitoring: NPM systems can ensure that adequate resources and priority levels are assigned to critical applications, reducing the chance of packet loss.
- Network Traffic Optimization: These systems can optimize network traffic to reduce the chance of packet loss.
- Alerts and Notifications: Proactive measures can be taken such as automatically rerouting traffic in case of network failure or congestion.
Packet loss is a common and impactful occurrence in computer networks. Understanding its causes, measurement techniques, and monitoring practices is essential for maintaining network performance. By implementing preventive measures and optimizing network configurations, administrators can minimize the amount of packet loss to ensure that the network is performing as needed to meet business requirements including ensuring employee productivity and customer satisfaction.
LiveAction’s LiveNX network performance management and LiveWire high-speed packet capture is the network monitoring solution of choice for organizations worldwide. LiveNX network monitoring software delivers network-wide observability, advanced analytics, and comprehensive reporting so you can optimize network and application performance everywhere. LiveWire generates high-performance telemetry for network and application performance analysis and real-time for detailed troubleshooting when existing data is not enough.
A recent Forrester Consulting study demonstrated that a 153% ROI was one of many benefits for companies that deployed LiveNX and LiveWire. Get in touch to learn more about network monitoring with LiveAction.
— by Susan Short. Susan is Director of Product Marketing for LiveAction.