Protocol Analysis and Troubleshooting

Expert-defined terms from the Professional Certificate in Network Performance Testing Techniques course at London School of Business and Administration. Free to read, free to share, paired with a professional course.

Protocol Analysis and Troubleshooting

ACK Storm #

ACK Storm

Definition #

A condition where a network device repeatedly sends acknowledgment packets, overwhelming the medium and causing latency spikes.

Example #

In a switched Ethernet, a misconfigured port sends ACKs for every frame, saturating the link.

Practical application #

Detecting an ACK storm helps isolate misbehaving NICs during performance testing.

Challenges #

Distinguishing between legitimate high‑volume ACK traffic and a storm requires baseline metrics.

ARP Spoofing #

ARP Spoofing

Definition #

The malicious insertion of false Address Resolution Protocol entries, redirecting traffic to an attacker’s device.

Example #

An attacker sends forged ARP replies, causing a victim’s traffic to route through the attacker’s machine.

Practical application #

Protocol analysis tools capture ARP packets to identify inconsistencies in MAC‑IP mappings.

Challenges #

Encrypted traffic can mask the impact, making detection reliant on statistical anomalies.

Asymmetric Routing #

Asymmetric Routing

Definition #

A scenario where forward and return traffic traverse different network paths, often leading to mismatched performance measurements.

Example #

A packet from Host A to Host B follows Path X, while the response travels via Path Y with different latency.

Practical application #

Troubleshoot by correlating timestamps from both directions to pinpoint divergent segments.

Challenges #

Correlating logs from disparate devices and handling NAT translations complicate analysis.

Baseline Performance #

Baseline Performance

Definition #

The measured performance characteristics of a network under normal operating conditions, used as a comparison point for troubleshooting.

Example #

Recording average latency of 12 ms on a 10 GbE link during off‑peak hours.

Practical application #

Establishing a baseline allows detection of deviations caused by configuration changes or faults.

Challenges #

Baselines must be periodically refreshed to account for evolving traffic patterns.

Bidirectional Flow #

Bidirectional Flow

Definition #

Traffic that includes both upstream and downstream packets within the same logical connection, often monitored to assess symmetry.

Example #

A TCP stream where client and server exchange data at similar rates.

Practical application #

Analyzing bidirectional flow helps verify QoS policies that guarantee equal bandwidth in both directions.

Challenges #

Asymmetric applications can skew measurements, requiring per‑flow normalization.

Buffer Overflow #

Buffer Overflow

Definition #

The condition where a network device’s buffer receives more data than it can store, leading to packet loss or device crash.

Example #

A router’s ingress queue fills during a traffic surge, causing subsequent packets to be dropped.

Practical application #

Monitoring queue depths during stress tests reveals buffer limits.

Challenges #

Differentiating intentional attacks from legitimate traffic spikes demands precise threshold settings.

CAPWAP #

CAPWAP

Definition #

Control And Provisioning of Wireless Access Points, a protocol that separates data and control planes for centralized Wi‑Fi management.

Example #

An AP receives configuration via CAPWAP from a controller, while user traffic bypasses the controller.

Practical application #

Analyzing CAPWAP control messages can uncover misconfigurations affecting throughput.

Challenges #

Encrypted CAPWAP tunnels obscure payload inspection, requiring decryption capabilities.

Checksum Validation #

Checksum Validation

Definition #

The process of verifying that a packet’s data has not been altered in transit by recomputing its checksum.

Example #

TCP checksum fails on a corrupted segment, prompting retransmission.

Practical application #

Tools that report checksum errors assist in diagnosing physical layer issues.

Challenges #

Offloading checksum calculation to NICs can hide errors from software‑based analyzers.

Coalescing #

Coalescing

Definition #

The technique of combining multiple small packets into a larger one to reduce CPU interrupts.

Example #

A NIC aggregates several TCP segments before delivering them to the OS.

Practical application #

Observing coalescing behavior helps explain lower interrupt rates during high‑throughput tests.

Challenges #

Coalescing can increase latency for latency‑sensitive applications, complicating performance interpretation.

Congestion Control #

Congestion Control

Definition #

Mechanisms that prevent network overload by adjusting transmission rates based on feedback.

Example #

TCP’s slow start and congestion avoidance phases adapt to perceived network capacity.

Practical application #

Evaluating how congestion control algorithms respond under simulated packet loss informs protocol robustness.

Challenges #

Different OS implementations may behave uniquely, requiring cross‑platform testing.

Cross‑Talk #

Cross‑Talk

Definition #

Unintended electromagnetic coupling between adjacent conductors, leading to signal degradation.

Example #

In a twisted‑pair cable, near‑end cross‑talk (NEXT) raises error rates on high‑frequency signals.

Practical application #

Measuring cross‑talk with a cable tester validates physical layer integrity before performance testing.

Challenges #

Environmental factors like temperature can affect cross‑talk levels, demanding repeated measurements.

DHCP Snooping #

DHCP Snooping

Definition #

A feature that filters DHCP messages based on trusted ports, preventing rogue servers from assigning IP addresses.

Example #

A switch blocks DHCP offers from an unauthorized device, preserving network integrity.

Practical application #

Analyzing DHCP snooping logs helps locate misbehaving hosts during troubleshooting.

Challenges #

Incorrect trust settings can inadvertently block legitimate DHCP traffic, causing widespread address allocation failures.

Disaggregation #

Disaggregation

Definition #

The practice of separating network functions into distinct hardware or software modules for independent testing.

Example #

Testing a virtual router’s forwarding plane separately from its control plane.

Practical application #

Disaggregated testing isolates performance bottlenecks to specific components.

Challenges #

Inter‑module communication overhead may mask true performance of individual parts.

Duplex Mismatch #

Duplex Mismatch

Definition #

A condition where two connected devices operate at different duplex settings, leading to collisions and reduced throughput.

Example #

One side set to full‑duplex while the other remains in half‑duplex, causing excessive retransmissions.

Practical application #

Detecting duplex mismatches through error counters and asymmetric throughput analysis.

Challenges #

Modern auto‑negotiation can obscure mismatches, requiring manual verification.

ECMP #

ECMP

Definition #

Equal‑Cost Multi‑Path routing, which distributes traffic across multiple paths of identical cost.

Example #

A router splits a flow across three parallel links to improve utilization.

Practical application #

Monitoring ECMP distribution helps verify that traffic is evenly balanced during performance tests.

Challenges #

Hashing algorithms can cause uneven flow distribution, especially with few large flows.

Flow Timeout #

Flow Timeout

Definition #

The period after which an inactive flow entry is removed from a device’s flow table.

Example #

A firewall deletes a flow after 30 seconds of inactivity, freeing resources.

Practical application #

Adjusting flow timeout values can affect latency and resource consumption during high‑volume testing.

Challenges #

Too short a timeout leads to excessive re‑creation overhead; too long retains stale entries.

Forwarding Plane #

Forwarding Plane

Definition #

The part of a network device responsible for moving packets based on forwarding tables.

Example #

A switch’s ASIC processes frames at line rate without CPU intervention.

Practical application #

Isolating forwarding plane performance isolates hardware limits from software processing delays.

Challenges #

Visibility into the forwarding plane is limited on proprietary ASICs, requiring vendor‑specific tools.

GRE Tunnel #

GRE Tunnel

Definition #

Generic Routing Encapsulation, a protocol that encapsulates a wide variety of network layer protocols inside virtual point‑to‑point links.

Example #

Routing IPv6 traffic over an IPv4 infrastructure using GRE.

Practical application #

Analyzing GRE encapsulation overhead helps assess impact on latency and MTU.

Challenges #

GRE does not provide encryption; combined use with IPsec adds complexity to troubleshooting.

HEP #

HEP

Definition #

High‑Efficiency Protocol, a method for aggregating multiple Ethernet links into a single logical connection while preserving low latency.

Example #

A data center spine uses HEP to combine four 25 GbE links into a 100 GbE logical pipe.

Practical application #

Measuring packet dispersion across aggregated links validates HEP’s load distribution.

Challenges #

Failure of a single physical lane can cause packet reordering, affecting latency‑sensitive applications.

ICMP Rate Limiting #

ICMP Rate Limiting

Definition #

The practice of limiting the number of Internet Control Message Protocol messages a device sends to prevent abuse.

Example #

A router caps echo‑reply messages to 100 per second.

Practical application #

Observing ICMP throttling helps explain missing ping responses during network stress tests.

Challenges #

Over‑aggressive rate limiting can mask legitimate network failures, complicating root‑cause analysis.

Ingress Policing #

Ingress Policing

Definition #

The enforcement of traffic contracts on incoming packets, dropping or marking packets that exceed defined limits.

Example #

A provider edge router discards packets that exceed a 10 Gbps inbound rate.

Practical application #

Monitoring ingress policing counters reveals whether performance degradation stems from policy enforcement.

Challenges #

Misconfigured policing can unintentionally penalize bursty traffic, leading to false‑positive congestion reports.

Jitter Buffer #

Jitter Buffer

Definition #

A storage area that temporarily holds incoming packets to compensate for delay variations before playback.

Example #

A VoIP endpoint uses a 20 ms jitter buffer to smooth out network jitter.

Practical application #

Adjusting jitter buffer size during testing can demonstrate trade‑offs between latency and packet loss.

Challenges #

Too large a buffer adds latency; too small results in audible glitches.

Definition #

A protocol that dynamically groups multiple physical links into a single logical link, providing redundancy and increased bandwidth.

Example #

Four 1 GbE ports combine into a 4 GbE LACP bundle.

Practical application #

Verifying LACP negotiation and load distribution ensures aggregated links behave as expected under load.

Challenges #

Mismatched LACP settings on endpoints can cause link flapping and traffic loss.

MAC Aging #

MAC Aging

Definition #

The process by which a switch removes inactive MAC address entries after a predefined period.

Example #

A switch clears a MAC entry after 300 seconds of inactivity, reverting to flooding for unknown destinations.

Practical application #

Observing MAC aging behavior helps diagnose intermittent connectivity issues.

Challenges #

Short aging timers can increase broadcast traffic, while long timers may retain stale entries after topology changes.

MTU Fragmentation #

MTU Fragmentation

Definition #

The division of a packet into smaller fragments to fit the Maximum Transmission Unit of a network segment.

Example #

An IP packet larger than 1500 bytes is fragmented when traversing a link with a smaller MTU.

Practical application #

Detecting excessive fragmentation indicates mismatched MTU settings or missing Path MTU Discovery.

Challenges #

Fragmented packets can be dropped by firewalls, leading to silent failures.

NetFlow Sampling #

NetFlow Sampling

Definition #

The practice of capturing a subset of packets for flow analysis to reduce processing overhead.

Example #

A router samples 1 out of every 1000 packets for NetFlow export.

Practical application #

Sampling rates affect accuracy of bandwidth utilization reports, requiring calibration during performance testing.

Challenges #

Low sampling rates may miss short‑lived flows, skewing traffic profiles.

NSL2TP #

NSL2TP

Definition #

Network Service Layer 2 Tunneling Protocol, a protocol that encapsulates Layer‑2 frames within IP for remote LAN extension.

Example #

Extending a VLAN across a WAN using NSL2TP tunnels.

Practical application #

Monitoring encapsulation overhead assists in evaluating end‑to‑end latency impact.

Challenges #

Interactions with firewalls and NAT devices can cause tunnel establishment failures.

Optical Power Budget #

Optical Power Budget

Definition #

The calculation of total allowable loss in an optical fiber link, balancing transmitter output and receiver sensitivity.

Example #

A 10 km fiber link with 0.5 dB/km attenuation and 2 dB connector loss fits within a 6 dB power budget.

Practical application #

Verifying power budget compliance prevents link failures during high‑speed testing.

Challenges #

Temperature variations affect attenuation, requiring dynamic adjustments.

Packet Capture Filtering #

Packet Capture Filtering

Definition #

The process of selecting specific packets for capture based on criteria such as protocol, address, or port.

Example #

Using a BPF expression “tcp port 80 and src host 10.0.0.5” to capture HTTP traffic from a client.

Practical application #

Focused captures reduce storage requirements and simplify analysis of targeted issues.

Challenges #

Overly restrictive filters may miss correlated packets, leading to incomplete diagnostics.

Path MTU Discovery (PMTUD) #

Path MTU Discovery (PMTUD)

Definition #

A mechanism where a sender discovers the smallest MTU along a path by sending packets with the “Don’t Fragment” flag set and reacting to ICMP “Fragmentation Needed” messages.

Example #

A TCP connection adjusts its segment size after receiving an ICMP “Fragmentation Needed” from an intermediate router.

Practical application #

Monitoring PMTUD failures reveals hidden MTU mismatches causing silent packet loss.

Challenges #

ICMP filtering by firewalls can break PMTUD, necessitating fallback strategies.

QoS Policing #

QoS Policing

Definition #

Enforcing bandwidth limits on traffic by dropping or remarking packets that exceed a configured rate.

Example #

A router enforces a 5 Mbps police on a video stream, discarding excess packets.

Practical application #

Evaluating the impact of policing on jitter and packet loss informs QoS policy tuning.

Challenges #

Policing can interact with upstream shaping, creating bursty traffic that confounds measurements.

RADIUS Accounting #

RADIUS Accounting

Definition #

The process of recording usage data from network devices for billing, auditing, or troubleshooting purposes.

Example #

A switch sends session start and stop records to a RADIUS server when a user connects.

Practical application #

Accounting logs help trace user‑initiated performance issues back to specific devices.

Challenges #

Inconsistent timestamps across devices can hinder correlation of events.

Round‑Trip Time (RTT) #

Round‑Trip Time (RTT)

Definition #

The elapsed time for a packet to travel from a source to a destination and back again.

Example #

An ICMP echo request and reply measured at 22 ms represent the RTT.

Practical application #

RTT measurements are fundamental to baseline latency and to detect path degradation.

Challenges #

Asymmetric routing can cause RTT to mask one‑way delay issues.

RST Injection #

RST Injection

Definition #

The act of sending forged TCP reset packets to terminate an existing connection.

Example #

An attacker injects an RST packet with the correct sequence number to disrupt a file transfer.

Practical application #

Detecting unexpected RSTs can indicate malicious activity or misconfigured devices.

Challenges #

Legitimate network events (e.g., NAT timeouts) can also generate RSTs, requiring contextual analysis.

SS7 Intercept #

SS7 Intercept

Definition #

Monitoring and potentially manipulating Signaling System 7 messages used in telephone networks.

Example #

Capturing an ISUP message to trace call setup pathways.

Practical application #

In telecom performance testing, SS7 intercepts reveal signaling delays affecting call quality.

Challenges #

Encryption and proprietary extensions limit visibility, demanding specialized equipment.

SFlow Sampling #

SFlow Sampling

Definition #

A technology that samples packets and exports flow information, typically using a fixed sampling rate.

Example #

A switch samples 1 out of every 4096 packets for sFlow export.

Practical application #

sFlow data provides real‑time visibility of high‑speed links where full capture is impractical.

Challenges #

Sampling introduces statistical variance; high‑precision analysis may require increasing the sample rate.

Spanning Tree Protocol (STP) #

Spanning Tree Protocol (STP)

Definition #

A protocol that creates a loop‑free topology by blocking redundant paths and reconverging after failures.

Example #

Rapid STP (RSTP) reconverges within 200 ms after a link failure.

Practical application #

Monitoring STP state changes helps locate topology‑related latency spikes.

Challenges #

Misconfigured priority values can cause suboptimal path selection, impacting performance.

TCP Fast Open (TFO) #

TCP Fast Open (TFO)

Definition #

An extension to TCP that allows data to be carried in the initial SYN packet, reducing connection setup latency.

Example #

A web browser sends an HTTP GET in the SYN during the first request to a TFO‑enabled server.

Practical application #

Measuring TFO effectiveness quantifies latency reduction for short‑lived connections.

Challenges #

Compatibility issues with middleboxes may cause TFO packets to be dropped, negating benefits.

Traffic Shaping #

Traffic Shaping

Definition #

The deliberate pacing of traffic to conform to a desired profile, often using token bucket algorithms.

Example #

A router shapes outbound traffic to a steady 2 Gbps rate, buffering bursts.

Practical application #

Shaping test traffic ensures consistent load on the device under test, improving repeatability.

Challenges #

Improper shaping can introduce additional latency and jitter, distorting performance results.

Definition #

Combining multiple upstream links to increase total outbound bandwidth and provide redundancy.

Example #

A data center edge router aggregates three 10 Gbps uplinks into a 30 Gbps pipe.

Practical application #

Verifying load distribution across aggregated uplinks validates redundancy mechanisms.

Challenges #

Uneven traffic distribution can cause some links to become bottlenecks while others remain underutilized.

VLAN Tagging #

VLAN Tagging

Definition #

The insertion of a VLAN identifier into Ethernet frames to separate traffic domains.

Example #

A frame carries a 0x10 VLAN tag to indicate the “Finance” VLAN.

Practical application #

Analyzing VLAN tags helps ensure proper segmentation and isolation during performance tests.

Challenges #

Mis‑tagged frames can be dropped or misrouted, leading to unexpected latency.

Virtual Extensible LAN (VXLAN) #

Virtual Extensible LAN (VXLAN)

Definition #

An overlay network technology that uses UDP to encapsulate Layer‑2 frames, enabling large‑scale segmentation.

Example #

A VXLAN network uses VNI 5000 to isolate tenant traffic.

Practical application #

Monitoring VXLAN encapsulation overhead informs decisions about overlay versus underlay performance.

Challenges #

Double encapsulation adds extra headers, increasing packet size and potentially causing MTU issues.

Wake‑On‑LAN (WoL) #

Wake‑On‑LAN (WoL)

Definition #

A mechanism that powers up a device remotely by sending a specially crafted Ethernet frame.

Example #

An admin sends a WoL packet to a dormant server to initiate a performance benchmark.

Practical application #

Understanding WoL timing helps schedule tests without manual intervention.

Challenges #

Network devices may block WoL packets, requiring configuration changes.

Weighted Random Early Detection (WRED) #

Weighted Random Early Detection (WRED)

Definition #

An algorithm that randomly drops packets before a queue becomes full, with weightings based on packet priority.

Example #

A router applies WRED to preferentially drop lower‑priority traffic under congestion.

Practical application #

Observing WRED behavior assists in tuning QoS policies to avoid tail‑drop scenarios.

Challenges #

Incorrect weight settings can cause high‑priority traffic to be unfairly penalized.

XML‑Based Configuration #

XML‑Based Configuration

Definition #

Storing device settings in XML format, often managed via network configuration protocols.

Example #

A router’s interface configuration is defined in an XML snippet transmitted via NETCONF.

Practical application #

Parsing XML configurations enables automated validation of settings before testing.

Challenges #

Complex schema versions can lead to parsing errors, requiring robust validation tools.

Zero‑Length Probe #

Zero‑Length Probe

Definition #

An ICMP echo request with no payload, used to verify reachability while minimizing traffic impact.

Example #

A monitoring system sends a zero‑length ping every 10 seconds to assess link health.

Practical application #

Zero‑length probes provide low‑overhead baseline latency measurements.

Challenges #

Some devices deprioritize or block zero‑length probes, skewing results.

802 #

1X Authentication

Definition #

A network access control protocol that requires devices to authenticate before gaining network connectivity.

Example #

A switch port denies traffic until the attached device successfully completes an EAP‑TLS exchange.

Practical application #

Capturing 802.1X EAPOL frames helps diagnose authentication failures that may appear as connectivity issues.

Challenges #

Misconfigured supplicants can cause repeated authentication attempts, inflating control traffic.

ARP Cache Poisoning #

ARP Cache Poisoning

Definition #

Manipulating a host’s ARP cache to associate an IP address with an attacker’s MAC address.

Example #

An attacker injects false ARP replies, causing a victim’s traffic to be redirected.

Practical application #

Detecting rapid ARP entry changes can signal a poisoning attempt.

Challenges #

High‑frequency legitimate ARP updates in dynamic environments can mask malicious activity.

Back‑Pressure #

Back‑Pressure

Definition #

A mechanism where a downstream device signals an upstream device to slow transmission, preventing buffer overflow.

Example #

A switch asserts pause frames to a sender when its egress queue exceeds a threshold.

Practical application #

Monitoring pause frame frequency helps assess whether back‑pressure is influencing latency.

Challenges #

Not all devices support Ethernet PAUSE, leading to asymmetric congestion handling.

Bandwidth Allocation #

Bandwidth Allocation

Definition #

The assignment of a specific portion of total bandwidth to a traffic class or user.

Example #

A router reserves 1 Gbps for critical VoIP traffic while allowing best‑effort traffic to share the remainder.

Practical application #

Verifying allocation policies during load testing ensures SLA compliance.

Challenges #

Over‑allocation can starve lower‑priority traffic, causing unexpected drops.

Bidirectional Forwarding Detection (BFD) #

Bidirectional Forwarding Detection (BFD)

Definition #

A lightweight protocol that provides rapid detection of path failures by exchanging short control packets.

Example #

Two routers exchange BFD packets every 10 ms, detecting a link loss within 30 ms.

Practical application #

BFD metrics aid in pinpointing failure detection latency in high‑availability designs.

Challenges #

Misconfigured timers can cause false alarms or delayed detection.

Bridge Loop #

Bridge Loop

Definition #

A network condition where two or more bridges create a circular path, causing endless frame replication.

Example #

Two switches connected without STP enabled flood broadcast traffic, saturating the network.

Practical application #

Detecting excessive broadcast frames signals a possible bridge loop.

Challenges #

Large Layer‑2 domains increase the difficulty of isolating the loop’s origin.

Checksum Offload #

Checksum Offload

Definition #

The delegation of checksum calculation to a network interface card, reducing CPU load.

Example #

A NIC computes TCP checksums for outgoing packets, freeing the host OS from that task.

Practical application #

Understanding offload behavior prevents misinterpretation of checksum errors in captured traffic.

Challenges #

Offload can hide errors from software analysis tools that expect software‑computed checksums.

Collision Domain #

Collision Domain

Definition #

A network segment where simultaneous transmissions can interfere with each other, leading to collisions.

Example #

All devices on a legacy hub share the same collision domain.

Practical application #

Reducing collision domains via switching improves throughput for performance testing.

Challenges #

Misplaced hubs in modern networks can still create hidden collision domains.

Congestion Notification (ECN) #

Congestion Notification (ECN)

Definition #

An explicit signaling mechanism where routers mark packets instead of dropping them to indicate congestion.

Example #

A router sets the ECN bits on a packet when its queue exceeds a threshold.

Practical application #

Observing ECN markings helps assess whether congestion control is being triggered without packet loss.

Challenges #

Not all end‑systems implement ECN, limiting its effectiveness.

Cross‑Connect #

Cross‑Connect

Definition #

A physical or logical connection that directly links two network ports without intervening switching.

Example #

A cross‑connect in a data center interconnects two routers for a dedicated high‑speed link.

Practical application #

Cross‑connects are used to isolate test traffic from regular production paths.

Challenges #

Managing cross‑connects manually can lead to configuration drift.

Deep Packet Inspection (DPI) #

Deep Packet Inspection (DPI)

Definition #

The analysis of packet payloads beyond header information to identify applications or threats.

Example #

A firewall uses DPI to block peer‑to‑peer traffic.

Practical application #

DPI logs assist in correlating performance degradation with specific application signatures.

Challenges #

Encryption limits DPI visibility, requiring TLS termination for full inspection.

Delay Variation #

Delay Variation

Definition #

The difference in latency between consecutive packets, often critical for real‑time services.

Example #

A video stream experiences 5 ms of delay variation, leading to occasional frame drops.

Practical application #

Measuring delay variation during tests helps assess suitability for VoIP or streaming.

Challenges #

Network congestion and route changes can cause sudden spikes in variation.

EtherChannel #

EtherChannel

Definition #

A Cisco‑specific implementation of link aggregation that bundles multiple physical links into a single logical channel.

Example #

Six 1 GbE interfaces form an EtherChannel with a combined bandwidth of 6 Gbps.

Practical application #

Verifying EtherChannel load balancing ensures that traffic is evenly distributed across member links.

Challenges #

Inconsistent hashing algorithms across devices can cause uneven distribution, impacting performance.

Forwarding Information Base (FIB) #

Forwarding Information Base (FIB)

Definition #

A data structure used by routers to make fast forwarding decisions based on the routing table.

Example #

A router’s ASIC consults the FIB to forward packets at line rate.

Practical application #

Examining FIB updates reveals route flaps that may cause temporary forwarding inconsistencies.

Challenges #

FIB size limitations can lead to route truncation in large networks.

GRE Tunneling Overhead #

GRE Tunneling Overhead

Definition #

The additional bytes added to each packet when encapsulated in a GRE tunnel, typically 24 bytes for header fields.

Example #

An IPv4 packet of 1500 bytes becomes 1524 bytes when GRE‑encapsulated.

Practical application #

Accounting for GRE overhead prevents inadvertent MTU violations during testing.

Challenges #

Multiple encapsulations (e.g., GRE over IPsec) compound overhead, complicating path MTU calculations.

Hash‑Based Load Balancing #

Hash‑Based Load Balancing

Definition #

A method that assigns packets to paths based on a hash of header fields such as source/destination IP and ports.

Example #

A switch uses a 5‑tuple hash to distribute traffic across four equal‑cost paths.

Practical application #

Understanding hash behavior helps predict whether a small set of heavy flows will fully utilize all paths.

Challenges #

Limited hash entropy can cause “hash collisions,” leading to uneven load.

IPsec Tunnel Mode #

IPsec Tunnel Mode

Definition #

A mode where the entire original IP packet is encapsulated and encrypted, providing confidentiality and integrity for the whole payload.

Example #

Two sites connect via an IPsec tunnel, each packet gaining a new outer IP header and ESP trailer.

Practical application #

Measuring added latency in tunnel mode quantifies cryptographic processing impact.

Challenges #

MTU reduction due to ESP overhead may cause fragmentation if not accounted for.

Jumbo Frames #

Jumbo Frames

Definition #

Ethernet frames with payloads larger than the standard 1500 bytes, commonly up to 9000 bytes.

Example #

A storage area network uses 9000‑byte jumbo frames to improve throughput.

Practical application #

Enabling jumbo frames across test equipment can reduce per‑packet processing overhead.

Challenges #

Inconsistent support across devices leads to frame drops or fragmentation.

Kernel Bypass #

Kernel Bypass

Definition #

Techniques that allow applications to interact directly with NIC hardware, bypassing the operating system kernel to achieve lower latency.

Example #

A high‑frequency trading application uses DPDK to achieve sub‑microsecond packet processing.

Practical application #

Evaluating kernel‑bypass performance provides a ceiling for achievable latency in optimized environments.

Challenges #

Requires specialized drivers and careful resource management to avoid packet loss.

Latency Budget #

Latency Budget

Definition #

The total allowable delay for a packet to travel through a network segment, often broken down into component contributions.

Example #

An industrial control system allocates 5 ms total latency, with 1 ms assigned to each hop.

Practical application #

Mapping latency budgets helps pinpoint which segment exceeds its allocation during testing.

Challenges #

Cumulative jitter can erode the budget even if individual hops meet their targets.

Link‑Layer Discovery Protocol (LLDP) #

Link‑Layer Discovery Protocol (LLDP)

Definition #

A vendor‑neutral protocol that allows devices to advertise identity, capabilities, and neighbor information.

Example #

A switch advertises its port ID, system name, and supported VLANs via LLDP.

Practical application #

LLDP data assists in building topology maps for troubleshooting.

Challenges #

Disabled LLDP on some devices can leave gaps in the discovered topology.

Multicast Snooping #

Multicast Snooping

Definition #

A switch feature that listens to multicast group membership messages to forward multicast traffic only to interested ports.

Example #

An IGMP query response informs the switch that only ports 3 and 5 should receive a particular video stream.

Practical application #

Verifying snooping tables ensures efficient multicast delivery during performance testing.

Challenges #

Incorrect snooping configuration can cause multicast storms or unintended traffic suppression.

Network Time Protocol (NTP) Synchronization #

Network Time Protocol (NTP) Synchronization

Definition #

The process of aligning device clocks to a common time source to ensure accurate timestamps across logs.

Example #

All routers synchronize to a stratum‑2 NTP server, maintaining sub‑millisecond accuracy.

Practical application #

Accurate timestamps enable precise correlation of events across multiple devices.

Challenges #

Network delays and jitter can affect NTP accuracy, requiring disciplined monitoring.

Optical Splitter Loss #

Optical Splitter Loss

Definition #

The attenuation introduced when an optical signal is divided among multiple fibers, typically expressed in decibels.

Example #

A 1:8 splitter adds approximately 9 dB loss to each output.

Practical application #

Calculating splitter loss is essential when designing passive optical networks for performance testing.

Challenges #

Manufacturing variations can cause actual loss to deviate from nominal values.

Packet Reordering #

Packet Reordering

Definition #

The phenomenon where packets arrive at the destination in a different order than they were sent, often due to parallel paths.

Example #

A load‑balanced link causes two packets to traverse different routes, arriving out of sequence.

Practical application #

Measuring reordering rates helps assess the impact on protocols that assume in‑order delivery.

Challenges #

High reordering can trigger unnecessary retransmissions, inflating perceived loss.

QoS Classification #

QoS Classification

Definition #

The process of assigning traffic to classes based on header fields, enabling differentiated handling.

Example #

A router classifies VoIP packets with DSCP EF for priority queuing.

Practical application #

Verifying classification rules ensures that test traffic receives the intended QoS treatment.

Challenges #

Overlapping rules can cause ambiguous classification, leading to unexpected queuing behavior.

Rate‑Based Scheduler #

Rate‑Based Scheduler

Definition #

A queuing algorithm that services queues at a fixed rate, regardless of packet size, to enforce bandwidth guarantees.

Example #

A scheduler allocates 5 Mbps to a video stream and 2 Mbps to a data backup flow.

Practical application #

Observing scheduler behavior reveals whether bandwidth guarantees are being met under load.

Challenges #

Burstiness can cause the scheduler to underutilize available capacity.

Spanning Tree Protocol (MSTP) #

Spanning Tree Protocol (MSTP)

Definition #

A variant of STP that supports multiple spanning tree instances, allowing per‑VLAN load balancing.

Example #

MSTP maps VLAN 10 to instance 1 and VLAN 20 to instance 2, each with its own root bridge.

Practical application #

Monitoring MSTP convergence times helps assess redundancy performance.

Challenges #

Incorrect MST region configuration can cause loops or suboptimal path selection.

TCP Congestion Window (cwnd) #

TCP Congestion Window (cwnd)

Definition #

A sender‑side variable that limits the amount of unacknowledged data in the network, influencing throughput.

Example #

During slow start, cwnd grows exponentially until packet loss is detected.

Practical application #

Tracking cwnd evolution during a test reveals how quickly a connection reaches full bandwidth.

Challenges #

Different OS TCP stacks implement cwnd growth differently, affecting cross‑platform comparisons.

Traffic Engineering (TE) #

Traffic Engineering (TE)

Definition #

The practice of directing traffic over specific paths to meet performance, cost, or policy objectives.

Example #

MPLS TE tunnels steer high‑priority traffic away from congested links.

Practical application #

TE

June 2026 intake · open enrolment
from £90 GBP
Enrol