Tele2 Speedtest Service
Connected directly to Tele2's international 100GE IP core network. Test your bandwidth with high-resolution latency, jitter, download, upload, and bufferbloat metrics.
HTTP Test Files Download
Raw uncompressed sparse test files hosted directly on Tele2 edge nodes. These files bypass disk I/O bottlenecks to saturate up to 10 Gbps of pure wire throughput.
HTTP Upload & FTP Service
Test upstream capacity directly through your browser or automate uploads using standard cURL and FTP protocols.
☁ In-Browser Upload Tester
Drag and drop any local file or generate a synthetic memory payload to measure your upstream speed in real-time.
Click to browse or drag & drop a file here
📁 FTP Server Specifications
High-performance `vsftpd` service configured on Tele2 core nodes for batch throughput benchmarking.
| Server Host | speedtest.tele2.net |
| Port | 21 (Standard FTP) |
| Authentication | Anonymous (any password) |
| Upload Directory | /upload |
| Cleanup Policy | Auto-purged on completion (cleandir) |
| Download Access | All sparse files mirrored in root |
Developer & SysAdmin Toolbox
Copy production-ready command lines for headless servers, Docker containers, remote SSH sessions, and CI/CD pipelines.
curl -o /dev/null -w "Speed: %{speed_download} B/s | Time: %{time_total}s\n" http://speedtest.tele2.net/100MB.zipTele2 Anycast Nodes
All requests to `speedtest.tele2.net` are automatically routed via BGP Anycast to the node with the lowest network latency.
Technical Details & Core Network
How the Tele2 Speedtest infrastructure sustains 10 Gbps+ per node without disk saturation.
🗜 Sparse Files & CPU-Bound I/O
Traditional file downloads are throttled by disk read speeds. Tele2 test files are generated as sparse files using UNIX zero-fill blocks:
dd if=/dev/zero of=sparse_10G.bin bs=1 count=0 seek=10G
The Linux filesystem allocates zero physical disk blocks for empty data. When Nginx reads the file, the kernel streams zeroes straight from CPU memory to the 10GE network interface at line rate (~1.06 GB/s).
🌐 BGP Anycast Routing
A single public IP address is announced simultaneously from multiple Tier-1 data centers across Europe via BGP Anycast.
Your ISP's routers calculate the shortest AS-path and lowest latency route, dynamically steering your traffic to the nearest Tele2 POP. If a node experiences maintenance, BGP automatically re-routes traffic to the secondary node with zero downtime.
🖥 Server Node Specifications
Each edge node runs on enterprise 1U Supermicro hardware:
- CPU: Octa-core Intel Xeon E3-1240 V2 @ 3.40 GHz
- Memory: 32 GB DDR3 ECC 1600 MHz
- Network Interface: Intel 10GE (82579LM chip with hardware offload)
- Operating System: Debian Linux with custom sysctl TCP tuning
⚙ TCP Optimization & Congestion Control
To eliminate packet buffer bottlenecks over high-latency cross-border links, TCP window scaling is enabled, HTTP gzip/brotli compression is explicitly disabled to avoid CPU throttling, and congestion control algorithms like BBR and CUBIC are utilized.
Research & Enterprise Inquiries
If you are an academic researcher, network engineer, or peering partner interested in high-performance network studies or dedicated 100GE measurements, please contact our network operations center.