Guides

Procedures for the tickets that keep coming back

Each guide starts from a situation a sysadmin will recognise — loss on a trace, a link that feels slow, a port already in use, a building that just got its power back — and ends with a result you can attach to the ticket. All of them assume a network you administer.

• Check which hosts came back after a power or switch outage

  Snapshot your authorised subnets before a maintenance window, rescan after the power or switch outage, and diff the lists to find hosts that stayed down.

  Uses Angry IP Scanner
• Find the hop where packet loss really starts — and when it's only ICMP deprioritisation

  A step-by-step method for locating the hop where packet loss begins, telling real loss from router rate limiting, and building evidence an ISP will accept.

  Uses PingPlotter
• Find which Windows process owns a port or a connection

  Identify the process behind a listening port or unexpected connection on Windows with TCPView, Tcpvcon, netstat and PowerShell, plus the PID traps.

  Uses TCPView
• Measure real LAN and WAN throughput with iPerf3

  How to run iPerf3 server and client tests, read TCP and UDP results, test both directions, and avoid the mistakes that make a 10G link look like 900 Mbit/s.

  Uses iPerf3