Hong Kong CN2 Bandwidth Testing Methods And Practical Tools For Quickly Diagnosing Network Bottlenecks

2026-06-08 21:57:28
Current Location： Blog > Hong Kong CN2
香港CN2

1.

Why conduct bandwidth and link tests for Hong Kong CN2

• CN2 (China Telecom Next Carrying Network) is a high-quality backbone network for international exports, suitable for services that require low-latency connections to users on the mainland.
• Testing can verify whether the bandwidth commitments (such as 100Mbps/200Mbps/1Gbps) are met, and identify link jitter and packet loss points.
• For servers, VPSs, hosts, or CDN nodes deployed in Hong Kong, bandwidth testing can guide DDoS protection and traffic scheduling strategies.
• Typical indicators: Bandwidth (Mbps), RTT (ms), packet loss rate (%), jitter (ms), and TCP handshake timing.
• It is recommended to repeat the tests at different times (during working hours, at night, on weekends) and with varying numbers of concurrent connections to evaluate stability.

2.

Common Bandwidth and Link Testing Methods (Commands and Instructions)

• iperf3: TCP/UDP performance benchmarks, recommended parameter examples: iperf3 -c server_ip -p 5201 -P 10 -t 30.
• speedtest-cli: HTTP/HTTPS speed testing, suitable for quickly verifying public network download/upload speeds: speedtest-cli --server SERVER_ID.
• ping / traceroute / mtr: Jump-by-jump latency and packet loss detection, mtr -r -c 100 target_IP can provide stable statistics.
• curl/wget download test: curl -o /dev/null -s -w "%{speed_download}\n" http://yourserver/file.bin is used for HTTP throughput.
• tcpdump / tshark: Packet capture analysis of TCP handshake, retransmission, and window scaling issues, examples: tcpdump -i eth0 -w trace.pcap host TARGET_IP.

3.

Practical Tools and Detailed Explanation of Common Parameters

• iperf3 Parameter Explanation: -P simulates the number of concurrent streams, -w adjusts the TCP window to avoid single-stream limitations, and -t sets the test duration to determine result stability.
• ethtool: View NIC capabilities and enable offload, example: ethtool -k eth0 ； It is possible to confirm whether GRO/TSO is enabled.
• sysctl Network Kernel Tuning: net.core.rmem_max, net.core.wmem_max, net.ipv4.tcp_Tools like rmem can improve throughput per connection.
• BBR and Congestion Control: Check and enable: sysctl net.ipv4.tcp_congestion_control=bbr ； It can significantly improve TCP throughput over long-distance links.
• iftop / nload: Real-time traffic monitoring to quickly identify processes or connections that are consuming bandwidth instantly.

4.

Steps and checklist for quickly diagnosing network bottlenecks

• Step 1: Verify whether the physical link and NIC are saturated (using ifconfig/ethtool/iftop).
• Step 2: Single-stream vs multi-stream testing (iperf3 -P 1 and -P 10) to determine if it is limited by the single connection window.
• Step 3: Track the path (mtr/traceroute) to determine whether packet loss/high latency occurs at the local data center or in the upstream backbone.
• Step 4: Check whether CPU/interrupt and kernel settings (top, sar, cat /proc/interrupts) are causing processing bottlenecks.
• Step 5: Regarding CDN and DDoS: Check for sudden traffic spikes or attacks, enable CDN caching and scrubbing nodes, and adjust ACLs and traffic thresholds.

5.

Example test data (Hong Kong CN2 VPS multi-point bandwidth test)

Test Location Download Mbps Upload Mbps Average latency in ms Packet Loss Rate %
Shenzhen, China (CN) 180 95 12 0.2
Guangzhou (CN) 165 90 14 0.3
Taiwan (TW) 220 110 8 0.1
West Coast of the US 80 40 120 0.5
• The table above shows the concurrent throughput measured using iperf3 -P10 -t30 at different test points for the same Hong Kong CN2 VPS. It can be seen that it performs better over the mainland link.
• If single-stream testing only reaches 60Mbps while multi-stream reaches 180Mbps, it is usually due to the TCP window or packet loss that limits a single connection.
• The combination of latency and packet loss determines the user experience, especially for real-time applications such as voice/video, where jitter is a concern.
• It is recommended to use the above tests as a baseline, and automatically collect data and trigger alerts on a regular basis (daily/weekly).

6.

Real cases: Hong Kong CN2 VPS Performance Improvement and Server Configuration Examples

• Background: An e-commerce company rented a CN2 VPS in Hong Kong. Users complained of order placement delays and slow page responses during peak hours in mainland China.
• Initial server configuration: 4 vCPU (Intel Xeon), 8GB RAM, NVMe 100GB, 1Gbps public network port, Ubuntu 20.04, kernel 4.19, BBR not enabled.
• Troubleshooting steps: Use mtr to locate that packet loss occurs 4 hops upstream ； iperf3 achieves only 55Mbps in single stream, and up to 160Mbps in multi-stream ； iftop shows it is not saturated.
• Take measures: Upgrade the kernel to 5.x and enable BBR (sysctl net.ipv4.tcp_congestion_control=bbr), adjust net.core.rmem_max/wmem_Set max and enable TSO/GRO, adjust MTU to 9000 (supported for intra-datacenter links).
• Results: The single-stream speed increased from 55Mbps to 180Mbps, average latency decreased by 20%, and page timeout issues were significantly reduced ； To combat low-volume DDoS attacks, add a CDN for static content acceleration and integrate with cloud scrubbing services.

7.

Practical recommendations for domain names/CDNs and DDoS protection

• Use a multi-node CDN (Hong Kong + mainland edge) for static content to reduce bandwidth strain on the origin server and lower latency.
• Set rate limits and connection count thresholds for important interfaces ； Work with WAF to defend against application-layer attacks.
• Negotiate minimum bandwidth and traffic billing terms with the bandwidth provider, and apply for a backup backhaul or off-site backup link if necessary.
• Deploy traffic scrubbing and blackhole strategies to adopt tiered responses against high-volume attacks in order to protect the control plane.
• It is recommended to set up a monitoring panel (Prometheus+Grafana) and alerting policies to cover metrics such as bandwidth, packet loss, latency, and the number of connections.

Tags：Hong Kong CN2 CN2 Bandwidth Testing VPS Bandwidth Testing Server Network Diagnosis DDoS Protection CDN Optimization Iperf3 BBR Routing Diagnosis More»
Previous article： Use Examples To Compare The Relationship Between The Price Of Hong Kong Servers CN2 And Actual Business Performance
Latest articles
Enterprises Expanding Markets To Sell Servers To Vietnam With Localized Pricing And After-sales System Setup
How To Test CN2 Japan Link Quality And Generate Visual Reports
Illustrated Guide To Setting Up IPs For Singapore Servers, Completing Network Segment Routing And Firewall Configuration
Key Points For Disaster Recovery Switching And Load Balancing Design For VPS Nodes At The Vietnamese Node In Enterprise-level Architectures
How To Determine How Much To Rent A VPS In Korea Based On Business Scale And Match Performance Requirements
Vietnamese CN2 Service Provider: Price And Service Comparison To Help You Choose Quickly
How Do Enterprises Assess The Time It Takes For Tencent Cloud Singapore Servers To Recover After A Failure?
Guidance On The Application Of Korean IP Native In SEO And Refined Promotion Operations
Cross-server StarCraft Battle, Creating A Room, Choosing A Korean Server, Multi-country Player Experience Analysis
Consider Multi-region Backups: Which Cloud Server In Taiwan Is Recommended With Excellent Disaster Recovery Capabilities?
Popular tags
American Sk High Defense Server Gaming Industry High-defense CDN Pictures Ordinary Vps Ssh Reinforcement Streaming Service Cn2gt Line Recommendation Server Speed Compliance United States VPS Login To Us Website Crawler Guide To Using VPS Small And Medium-sized Enterprises Vps High Protection Fault Recovery Automatic Kicking Host Technology American Services Defense Measures American VPS Hosting User Impact Us Cn2 Space Performance Comparison American Onion Server Business Impact Server Configuration Guide Website Building
Related Articles