In this lesson, you will learn how to look beyond simple ICMP "ping" requests to identify exactly where a packet’s path fails within your network stack and routing topology.
Beyond the Ping
While a standard ping confirms end-to-end connectivity, it is often blind to the specific hop causing latency or packet loss. In advanced network administration, we need to distinguish between a dead host, a misconfigured router, and a silent firewall drop. To achieve this, we use diagnostic tools that leverage the kernel’s path discovery mechanisms, specifically traceroute and mtr.
Advanced Path Diagnostics with mtr
mtr (My TraceRoute) combines the functionality of traceroute and ping by sending a continuous stream of packets to each hop along the path. By observing the statistics for every router in the chain, you can pinpoint exactly where the performance degradation or reachability issue originates.
When investigating a path, use the following command to see the real-time interaction with every hop:
mtr -rw google.com
-r: Puts the program into report mode.-w: Displays a wide output, providing space for full hostnames.
Interpreting Kernel Behavior
When you perform path diagnostics, you are effectively triggering the kernel to generate ICMP TIME_EXCEEDED messages or specific error codes from intermediate routers. If you see a hop showing 100% packet loss followed by successful hops, this does not necessarily mean the network is broken; it often indicates that the router in question is rate-limiting ICMP responses to protect its own control plane.
By monitoring the Loss% and Snt (Sent) columns, you can differentiate between a permanent routing failure (where 100% loss persists to the end) and administrative packet filtering. This visibility allows you to verify if your routing tables and policy-based routes are directing traffic through the intended path or if a tunnel—such as a VXLAN or GRE link—is silently dropping transit frames.
Key Takeaway: Use mtr to transform static path snapshots into dynamic reachability statistics, allowing you to isolate network faults to specific router hops rather than assuming a total path failure.