NSComm Giant 663 TAP Network Packet Broker | 32 x 100GbE Ports

Q1: What is a Network Packet Broker (NPB), and why do I need it?

A: A Network Packet Broker (NPB) is a traffic visibility platform that collects, aggregates, filters, replicates, and load-balances network packets from multiple links to monitoring and security tools. It solves common challenges such as tool overload, blind spots, and inefficient traffic duplication, ensuring the right traffic reaches the right tool at the right time-especially in environments with high-speed 25G/100G/400G networks.

Q2: What are the key differences between Giant 662, Giant 663, and Giant 674?

A: The three models are designed for different network scales and interface requirements:

• Giant 662: Best for mixed-speed deployments with 48 × 1/10/25G + 8 × 40/100G, ideal for 25G/100G hybrid networks.
  
• Giant 663: High-density 100G-focused platform with 32 × 40/100G, optimized for large data centers and carrier networks.
  
• Giant 674: Next-generation high-speed visibility with 24 × 40/100G + 8 × 100/400G, built for 400G-ready environments and future expansion.
  

Q3: Can the ports be configured as either input or output?

A: Yes. All three models support flexible port roles. Ports can be configured as input (ingress) or output (egress) depending on your architecture-aggregation, distribution, replication, or load-balanced delivery. This flexibility allows you to build many-to-one, one-to-many, or many-to-many traffic visibility designs without fixed-port constraints.

Q4: Do these NPBs support breakout (port split) configurations?

A: Yes. Breakout is supported across the product family:

• 100G ports can be broken into 4 × 25G or 4 × 10G (depending on deployment).
  
• Giant 674 also supports 400G → 4 × 100G breakout, enabling high-density expansion without additional hardware. Always ensure breakout cables and transceivers match the interface type (QSFP28 / QSFP-DD) and the target speed.
  

Q5: How do these devices ensure zero packet loss at full line rate?

A: These platforms are designed for full-duplex, line-speed processing, meaning they can handle traffic at full link speed while maintaining reliable forwarding and distribution. To achieve the best results in production networks, it's recommended to apply filtering policies and load balancing to avoid excessive replication or oversubscription of outputs, which is a common cause of packet drops in real-world deployments.

Q6: What monitoring tools can the Giant series feed traffic to?

A: The Giant series is designed to work with common security and performance tools such as IDS/IPS, NDR, SIEM, APM, NPMD, packet analyzers, signaling analyzers, and forensic platforms. It supports traffic replication and multi-output distribution, so one link can feed multiple tools simultaneously-while filtering ensures each tool receives only what it needs.

Q7: Can I replicate one traffic stream to multiple tools at the same time?

A: Yes. All three models support one-to-many traffic replication. You can take traffic from one or multiple input interfaces, apply filters or classification rules, and output to multiple monitoring tools simultaneously. This is commonly used in SOC architectures where IDS/IPS, SIEM collectors, and forensics tools all require parallel access to the same traffic.

Q8: Can it load-balance traffic across multiple monitoring tools?

A: Yes. Load-balanced distribution is supported to scale monitoring capacity. When traffic volume exceeds the capability of one tool, the NPB can distribute flows across multiple output ports using classification rules such as five-tuple-based strategies, helping maintain tool performance and preventing drops caused by tool overload.

Q9: What filtering and traffic classification methods are supported?

A: The Giant series supports packet classification based on:

• Input interface
  
• Five-tuple (src/dst IP, src/dst port, protocol)
  
• Fixed feature fields / characteristic matching
  

This enables precise traffic selection and policy control, reducing tool noise and allowing per-tool traffic shaping according to operational needs.

Q10: Does the Giant series support tunneled traffic (GTP / GRE / VXLAN)?

A: Yes-tunnel awareness is included in the family.

• Giant 662 supports GTP / GRE / VXLAN, and can steer based on internal traffic characteristics.
  
• Giant 663 and Giant 674 support GTP / GRE and can distribute traffic based on inner-layer IP addresses, which is critical in telecom and overlay network environments. This helps visibility teams analyze real traffic sessions rather than only outer tunnel headers.

Q11: Can I strip tunnel headers or steer traffic by inner IP?

A: Yes. The devices support tunnel-based traffic handling functions such as internal IP diversion and tunnel header stripping (depending on configuration and policy). This capability helps monitoring tools receive packets in a cleaner format and enables accurate analytics on tunneled traffic.

Q12: Do these models support packet slicing? Why is it useful?

A: Yes. Packet slicing is supported (configurable by specified length). This is useful when you want to reduce monitoring bandwidth, storage requirements, or tool CPU usage by delivering only the packet headers or the first N bytes. Packet slicing is commonly used for performance monitoring, metadata analysis, and compliance scenarios where payload is not required.

Q13: What management options and APIs are available?

A: Management interfaces include:

• Web UI (visual management and configuration)
  
• CLI (local/remote command-line control)
  
• SNMP (monitoring and integration with NMS platforms)
  
• RESTful API / RPC interfaces (automation and integration into network operations workflows)
  

These options support both day-to-day operations and large-scale deployments requiring automation.

Q14: What power supply options are supported (AC/DC redundancy)?

A: The Giant series supports redundant power supply options for high availability.

• Giant 662 and Giant 663 support redundant AC and optional DC variants (depending on model).
  
• Giant 674 supports redundant AC and DC options and is designed for high-density deployments.
  

For telecom environments, DC variants typically support -36V to -72VDC ranges, suitable for carrier-grade power systems.

Q15: What are the most common deployment mistakes-and how can they be avoided?

A: The most common issues in real projects include:

1. Oversubscription: Sending more traffic to outputs than the tool or link can handle. Use filtering and load balancing.
   
2. Uncontrolled replication: Replicating "all traffic to all tools" leads to tool overload. Apply per-tool policies.
   
3. 
   Breakout mismatch: Incorrect cables/transceivers cause link failures. Ensure correct QSFP/QSFP-DD breakout type and speed.