VPS CPU spikes and missed exits: monitoring worth setting up

[FXS London Scalper]

A VPS keeps your platform running, but VPS problems are often invisible until they cause a missed exit. High CPU, low memory, or network issues can delay trade management.

Common causes of trouble:
- Too many charts and indicators open.
- Multiple platforms on a small VPS.
- Windows updates running during trading hours.
- Memory leaks in custom indicators or EAs.

Monitoring to set up:

1. CPU and memory alerts: a free tool can send a notification if CPU stays above 80% for more than a minute.
2. Platform ping to the trade server: MT5 shows this in the bottom right corner. Log it periodically.
3. Windows update scheduling: set active hours to cover your trading sessions.
4. Regular restarts: weekly restart on the weekend clears memory leaks.

The easiest fix for most issues is to run fewer things. A scalping VPS needs one platform, a few charts and the EA. Everything else can run elsewhere.

What monitoring do you have on your VPS? Has a spike ever cost you a trade?

[Shadow Trader]

A scalper leaves a platform running on a laptop. The screen dims, the power plan kicks in, and the network adapter drops into a low-power mode. The platform stays open, but the connection stalls for a second or two every so often, and nobody connects it with the power settings.

Power-saving options on network adapters and USB devices are a well-known cause of strange latency spikes. They're switched on by default on many laptops and rarely noticed until a trade goes wrong.

Switching to a high-performance power plan, disabling "allow the computer to turn off this device to save power" on the network adapter, and keeping the machine plugged in fixes most of it. Wi-Fi adapters have a similar setting, often labelled power saving mode, worth checking at the same time.

Has any hidden setting like that turned out to be the cause of a latency problem?