Scale Up and Lock Down Your Networks With Network Traffic Generators
Key takeaways:
- Network traffic generators enable simultaneous testing of functionality, scalability, performance, and security throughout the development lifecycle.
- Both hardware and software traffic generators are available. Though they can be used independently, they are often used together because they complement each other's focus areas.
- The ability to automate traffic generators is essential for properly testing complex network scenarios and protocol sequences.
Hardware and software engineering teams often focus more on making their devices or services work correctly for a single user but tend to ignore the scaling, performance, and security problems that may emerge when thousands of users try their products on release day. Ignoring the latter may result in embarrassing failures under a public spotlight, customer churn, and reputational damage.
Is there a systematic way to force scalability and security into your daily stand-up updates? The good news is yes! Hardware and softwarenetwork traffic generators enable your entire infrastructure to be continuously tested for network-related scalability, functionality, and security problems.
In this blog, explore how network traffic generators are being used to accelerate and protect even the largest complex networks.
What is the purpose of a network traffic generator?
Fig 1. Keysight IxNetwork network traffic generator scenario builder
Network traffic generators synthesize network and application traffic that is realistic in content and volume, send that traffic through a target network, and help engineers assess:
- all the network devices for correct functionality
- network stress testing metrics, like load time, bandwidth, jitter, packet loss, and delay; quality of service (QoS); and quality of experience (QoE)
- protocol load testing metrics, like the line rate and optimal packet size
- application performance metrics, like the number of requests served per second or the number of dropped video call packets
How do network traffic generators assist in testing and validating network performance?
Network traffic generators facilitate many comprehensive and robust network testing and validation as outlined below.
1. Realistic network and application traffic synthesis
Traffic generators can create a wide range of traffic types and profiles to test networks using realistic simulations of user behaviors and application interactions.
They can emulate traffic profiles observed in data centers, corporate networks, and public networks.
They can also emulate niche use cases. For example, mobile networkpacket generators can synthesize protocols like service data adaptation protocol (SDAP) or radio resource control protocol (RRC) that are useful for testing 4G/5G/6G networks. Additionally, they can model phenomena like dynamic subscriber behavior and multiple application services per user, ensuring comprehensive and genuine network traffic scenarios.
2. Scalable high-performance traffic generation
Network traffic generators can synthesize high volumes of traffic, terabytes of data, and millions of traffic flows simultaneously for testing high-capacity networks, routers, and switches.
- With the help of dedicated traffic generator hardware like Keysight's AresONE and UHD400T, these tools are capable of terabit-scale simulations with millions of concurrent traffic streams that mimic users as well as application programming interface (API) clients.
3. Comprehensive list of supported protocols and network services
Good network traffic generators support a wide range of traffic types, protocols, and payloads to simulate a realistic internet mix (or IMIX), which refers to typical real-world traffic patterns.
Traffic generators can simulate a variety of common protocols like user datagram protocol (UDP), transmission control protocol (TCP), internet protocol (IP, IPv4, or IPv6), address resolution protocol (ARP), internet control message protocol (ICMP), transport layer security (TLS), and many more.
They can simulate different traffic types like:
- multi-protocol label switching (MPLS)
- Ethernet traffic
- virtual local area network (VLAN) segmenting
- IPsec
They also enable extensive testing of various application scenarios, including:
- web services
- video streaming
- voice-over-IP (VoIP)
- data transfer protocols
- storage services
4. Dynamic testing with detailed analysis
Traffic generators allow changing network characteristics — like the frame rate, frame size, and packet header fields — without stopping a test. This adds a lot of flexibility for simulating diverse network conditions and changes.
Detailed flow tracking, real-time filtering and sorting, flow visualizations, and other analyzer features allow in-depth analysis of metrics like loss, latency, jitter, and errors, which is essential for performance testing.
5. Advanced automation
Good generators provide extensive support for automating their traffic simulation engines. This facilitates repeatable and reproducible tests; the ability to set up complex network topologies, conditions, and changes; and integration with automated testing pipelines for agile and DevOps environments. The automation can either be through easy-to-use graphical user interfaces (GUIs) or through APIs to be used by scripts written in programming languages (like Python).
6. Benchmarking, error injection, and specific use case testing
Network generator tools conform with common benchmarking request-for-comment (RFC) pstandards like:
- the benchmarking methodology for network interconnect devices (RFC 2544)
- the benchmarking methodology for LAN switching devices (RFC 2889)
- the methodology for IP multicastbenchmarking (RFC 3918)
Generators also support error injection, like malformed packets, to validate network robustness.
Specific use case coverage includes testing multicast traffic, carrier Ethernet, software-defined wide area network (WAN) deployments, and network security.
What traffic patterns and scenarios can be generated using network traffic generators?
Fig 2. IxNetwork test solution for software-defined core network and data center network
The list below, while far from comprehensive, gives an idea of the scenarios that traffic generators can simulate:
- Hybrid cloud data centers: Traffic generators are used to benchmark virtual and cloud-native infrastructure performance by simulating realistic data center workloads.
- Multicast traffic scenarios: Network traffic generators allow qualification and performance testing of multicast services like internet protocol television (IPTV).
- Video and voice testing: Network traffic generators like Keysight's IxLoad can validate the throughput and quality of voice services on network services like VoIP and voice-over-long-term-evolution (VoLTE) by synthesizing voices and tones. It can also emulate thousands of interactive live-streaming user sessions and measure video quality in real time. Different video formats and adaptive bitrate streaming protocols are supported for realistic load scenarios by generating video streams with different codecs and formats to accurately replicate real-world conditions.
- Ethernet testing: Carrier Ethernet services and advanced features, like time synchronization in automotive and industrial Ethernet, can be tested with the help of network traffic generators.
- Advanced network topologies: Traffic generators enable comprehensive testing of complex routing and switching configurations, load-balancing configurations, MPLS, software-defined networks (SDNs), software-defined WANs (SD-WANs), data center topologies, public broadband networks, and more.
How does a network traffic generator simulate real-world conditions such as congestion and packet loss?
Fig 3. IxLoad emulates realistic user behaviors
For simulating real-world conditions, traffic generators use different approaches like:
- Adaptive control techniques: The testing tool simply cranks up its synthesized traffic mix and volume until it observes impaired conditions or elevated error rates.
- Statistical and machine learning models: These techniques learn probabilistic models from real network traffic captured using tools like Wireshark. The models enable them to identify the conditions that will probably lead to impaired conditions like network congestion or packet loss. The traffic generator can then recreate the combination of conditions identified by the models to increase the chances of congestion and packet loss.
- Replay captured traffic streams: Another approach is to capture actual network traffic during real-world operations and replay them later during synthesis.
- Packet corruption: For simulating packet losses, a traffic generator may intentionally inject errors into its packets, knowing that some network device downstream will reject them.
- Manipulating packet delays: Manipulating jitter (the variability of packet delays) to make packets arrive out of order forces them to be dropped in many scenarios, but not all. Some services tolerate out-of-order packets.
- Special algorithms: Traffic generators like IxLoad also use special algorithms like dynamic control plane (DCP) and network failure threshold (NFT). DCP dynamically creates a new interface in sync with a new subscriber during testing and tears down the interfaces later. This replicates real-world subscriber behavior. NFT enables the generator to run tests even when some interfaces fail during the test run until the number of failures crosses a threshold.
How do network traffic generators help assess the scalability of network infrastructure?
Hardware traffic generators test network scalability by:
- synthesizing millions of trackable flows
- configuring thousands of flow groups, each with a unique transmit profile
- supporting thousands of hardware streams per port
- binding to millions of MPLS labels
- utilizing their high port density to validate the performance and scalability of high-speed optical networking devices that support current and next-generation speeds with state-of-the-art modulation like phase-amplitude-modulation (PAM4)
- emulating congestion in storage network workloads involving remote direct memory access (RDMA) over converged Ethernet (RoCEv2) and non-volatile memory express over fabrics (NVMe-oF) interfaces
How do network traffic generators contribute to security testing?
Network traffic generators are extensively used for testing security appliances, intrusion detection, and other cyber attack identification. These capabilities include testing:
- Firewalls: They generate malicious and valid traffic for testing firewall effectiveness.
- Authentication and authorization: Traffic generators can be configured to test these security aspects as well as transmit the right credentials to gain access to protected functionalities.
- IPsec and virtual private networks (VPNs): Generators support the emulation of IPsec tunnels and can generate application traffic over these encrypted tunnels. They also test the performance and security of VPNs.
- Transport layer security: Generators emulate clients and servers using different versions of TLS and different combinations of cipher suite negotiations to validate the performance and security of network devices handling encrypted traffic.
- MACsec encryption: Generators can benchmark MACsec — a standard for data confidentiality and integrity at the media access (MAC) layer for any MAC address — encryption performance under real-world cloud and data center workloads.
- Network and application security: Generators validate network security tools by generating realistic traffic, threats, and application behaviors to measure performance and security efficacy.
Keysight network traffic generator products
Fig 4. Keysight network traffic generator hardware products
Keysight's high-capacity hardware traffic generators include:
- AresONE: The AresONE offerings can generate traffic for gigabit Ethernet bandwidths ranging from 10 to 800 gigabits-per-second.
- UHD400T: The UHD400T is a high-density 400GE traffic generator with features like dynamic control plane scaling.
- Artificial intelligence (AI) data center test platform: This test platform is an 800GE/400GE traffic generator solution that can realistically emulate modern AI workloads.
- PerfectStorm: PerfectStorm can simulate the traffic of millions of real-world end-user environments and their applications.
- CloudStorm: CloudStorm is a module that allows concurrent emulation of complex application traffic, unprecedented TLS-encrypted applications, and a large volume of distributed denial-of-service (DDoS) traffic to validate that your network infrastructure is highly scalable and secure.
- Novus: Novus load modules are based on IxOS hardware and support the validation at all layers using IxNetwork and IxLoad software.
Fig 5. Keysight network traffic generator software products
The software-based traffic generator tools include:
- IxNetwork: IxNetwork is a layer 2-3 generator that can mimic realistic user applications and scenarios with comprehensive coverage for routing, switching, MPLS, industrial Ethernet, data center networking, and SDNs.
- IxLoad: IxLoad is a layer 4-7 generator for performance testing multiplay services, application delivery platforms, and network security appliances.
- IxChariot: IxChariot uses distributed software agents to generate real-world application traffic loads, collects performance metrics centrally, and features dashboards for holistic data analysis.
- Ixia-c: This is a containerized softwaretraffic generator that can synthesize complex data plane traffic and implements the opentraffic generatorAPI.
- Elastic network generator: The Elastic Network Generator is a write-once-run-anywhere abstraction solution over diverse generators like Ixia-c, UHD400T, and Novus.
- Fabric emulator: Fabricemulator can mimic complex data center topologies with realistic traffic constraint conditions.
- CyPerf: CyPerf allows network equipment manufacturers to validate their products by replicating real-world traffic that mirrors their customer’s actual deployments.
- Eggplant performance: Eggplant performance acts as both server and client to generate traffic loads to test websites and applications, network infrastructures, intermediate layers of N-tier architectures, and back-end equipment like routers and content-aware network devices.
- BreakingPoint: BreakingPoint is a security-testing software that can generate real-world loads including legitimate traffic, DDoS, and malware.
- TrafficREWIND: TrafficREWIND captures traffic from production networks and converts them into test traffic profiles that can be imported into BreakingPoint.
Boost your infrastructure performance with Keysight network traffic generators
In this article, we introduced network traffic generators and their importance in testing network functionality and performance.
Keysight offers a comprehensive set of hardware and software traffic generators for every use case. Contact us for advice on selecting the best network traffic generators and other network testing tools for your organization and service.