Finding where the problem is.
WHAT TO ESTABLISH FIRST
Whether the problem affects one destination or all Whether it affects one network or several Whether it is constant or time-dependent
WHAT A TRACE SHOWS
The path taken, and latency at each hop.
WHAT IT DOES NOT SHOW RELIABLY
The return path, which may differ entirely.
WHY THAT MATTERS
Asymmetric routing means the problem may be on a path your trace never displays.
WHAT TO BE CAREFUL INTERPRETING
High latency at one hop that does not persist at later hops.
WHY
Routers deprioritise responding to traces, so an intermediate hop's figure may mean nothing.
WHAT ACTUALLY INDICATES A PROBLEM
Latency or loss that persists at every subsequent hop.
WHAT TO USE FOR ONGOING MEASUREMENT
Tools combining trace and continuous probing, showing loss and latency per hop over time.
WHAT TO TEST FROM
Several locations and networks, since the problem may be specific to one path.
WHAT LOOKING GLASSES PROVIDE
The view from another network, showing how they reach you.
WHAT TO RECORD
Times, paths, and measurements, when reporting to a provider.
WHY
Without them, the report cannot be investigated.
