A server migration changed the fill speed and nobody was told

[Shadow Trader]

One month a scalper's average fill time is 45 milliseconds. The next month it's 140. Strategy unchanged, VPS unchanged, spreads normal. Results get worse and the cause isn't obvious.

Eventually the answer turns up in a forum thread. The broker moved its trade servers to a different data centre, or switched liquidity bridge software, or changed routing. Nobody was notified because from the broker's side nothing relevant had changed.

For a scalper, the effect is real. A ping that used to be 3 milliseconds is now 60, because the VPS sits in the old data centre and the broker is in the new one.

Several signs point to a migration. A sudden change in ping to the trade server. A change in the server name or IP address shown in the platform. A shift in the pattern of slippage. A new platform build number.

Keeping a record of ping and average fill time each week turns a mysterious decline into an obvious event with a date on it.

With that evidence, asking the broker directly is productive, and in some cases the VPS can be moved to match.

Has a drop in results ever been traced to a change on the broker's side?

[FXS Prop Desk]

Keeping a record of ping and average fill time each week turns a mysterious decline into an obvious event with a date on it.

A jump from 45 to 140 ms with nothing changed on the client side is the kind of thing a monthly fill-time average catches and gut feeling doesn't. Keeping that monthly figure next to the broker's announcements page also turns a vague complaint into a dated, specific question support has to answer: what changed on the server side in that week?