Route changes: the day the ping tripled with no explanation
Posted: Sat Oct 31, 2026 11:46 pm
Connection speed that was fine for months suddenly gets worse. Ping to the broker goes from 20 ms to 60 ms. Nothing changed locally. The cause is often an upstream routing change.
Internet traffic takes a path through many networks. Providers adjust routing for cost reasons, congestion, or maintenance, and a path that used to go direct may start going through a distant exchange. The user sees only the result.
A traceroute shows each hop and the time at each. Comparing a traceroute from a good day with one from a bad day usually reveals where the extra delay appeared. Sometimes it's the home ISP, sometimes a transit provider in the middle, and sometimes the broker's own network.
Reporting it with evidence tends to work better than a general complaint. ISPs and VPS providers can sometimes reroute or escalate when handed a clear trace.
Regularly logging ping results to the broker, say every few minutes in the background, builds a baseline. Then a change in routing stands out right away rather than lingering unnoticed for weeks.
Another trick is testing two or three alternative servers from the same broker, if available. Many brokers run several trading servers, and the ping to each differs. A quick test of each, repeated at different hours, may find a server that is consistently better from the trader's location. Switching costs nothing but a few minutes.
Who noticed a sudden jump in latency first, the trader or the monitoring script?
Internet traffic takes a path through many networks. Providers adjust routing for cost reasons, congestion, or maintenance, and a path that used to go direct may start going through a distant exchange. The user sees only the result.
A traceroute shows each hop and the time at each. Comparing a traceroute from a good day with one from a bad day usually reveals where the extra delay appeared. Sometimes it's the home ISP, sometimes a transit provider in the middle, and sometimes the broker's own network.
Reporting it with evidence tends to work better than a general complaint. ISPs and VPS providers can sometimes reroute or escalate when handed a clear trace.
Regularly logging ping results to the broker, say every few minutes in the background, builds a baseline. Then a change in routing stands out right away rather than lingering unnoticed for weeks.
Another trick is testing two or three alternative servers from the same broker, if available. Many brokers run several trading servers, and the ping to each differs. A quick test of each, repeated at different hours, may find a server that is consistently better from the trader's location. Switching costs nothing but a few minutes.
Who noticed a sudden jump in latency first, the trader or the monitoring script?