Ping, traceroute and MTR live from the KernelHost backbone in Maincubes FRA01. Plus 100 MB and 1 GB speedtest files. Checked against private addresses, protected by hCaptcha.
A Looking Glass is a publicly accessible network diagnostic tool provided by a carrier or data center. It runs commands like ping, traceroute or MTR from its own backbone, so external users can see how a target behaves from the perspective of that specific location. KernelHost runs this Looking Glass from Frankfurt am Main, hosted in Maincubes FRA01 with peering at DE-CIX, the largest internet exchange in Europe.
Two test files sit directly on this server: 100 MB for a quick bandwidth measurement, 1 GB for a stable measurement over several seconds. Both contain random data and are not compressible, so the result reflects the real bandwidth between Frankfurt and your connection. Note: downloads are throttled to 5 per hour and IP to prevent bandwidth abuse.
The Looking Glass server is located in Maincubes FRA01 (Offenbach am Main) with redundant connectivity via NTT Global Data Centers Frankfurt 1. Direct access to DE-CIX, native IPv4 and IPv6, core transit through multiple Tier-1 providers. The results shown here reflect the path between exactly this location and your target.
A Looking Glass is a public network tool provided by carriers and data centers. It runs diagnostic commands like ping, traceroute or MTR from its own backbone, so external users can check how their network behaves from the perspective of that location.
Frankfurt is the largest internet exchange in Europe (DE-CIX). A ping from the KernelHost backbone in Frankfurt shows latency and packet loss between FRA01 and your target. That helps when comparing routes, hunting down network issues, and making decisions about location and connectivity.
Traceroute lists every router (hop) on the path to the target along with its latency. This makes it easy to spot where a connection slows down or breaks. Private or reserved address space is not queried here: the tool checks the target up front.
MTR (My Traceroute) combines ping and traceroute. It continuously sends packets to each hop and shows loss and latency per hop in a table. This is ideal for finding sporadic problems along the route, instead of seeing only a single snapshot.
Intentionally so. The tool blocks RFC1918, loopback, link-local, CGNAT and reserved address space. Otherwise any anonymous visitor could enumerate internal KernelHost hosts or third-party private networks from our backbone (SSRF class). Tools like this need exactly that anti-SSRF layer.
Two test files (100 MB and 1 GB) sit directly on this server. They let you measure download bandwidth from the KernelHost backbone in Frankfurt to your connection. The files contain random data and are not compressible, so the result reflects real bandwidth.
No. There's deliberately no API. Every request requires hCaptcha plus IP-based rate limits (20 tool commands per minute, 5 speedtest downloads per hour). This protects both our backbone resources and the broader internet from abuse of this host.
Hostnames and IPs you enter are used only for resolution and execution of the command. There's no long-term storage of inputs or outputs, no tracking, no profiling. Standard Apache logs (IP, path, timestamp) are kept short-term for abuse mitigation.
Need more than just tools? Take a look at our commercial hosting lineup.