Skip to main content
Blog/VPS

MT5 VPS latency test: what to measure

A low-latency VPS is not automatically a reliable trading setup. You need to test distance to the broker, terminal stability and whether the account keeps reporting after deployment.

Infrastructure6 min readJul 3, 2026
By xTriel editorial team · Reviewed and updated
Measure the full execution path: Terminal ping, Order round trip, and Stability.
FIG 01 — Measure the full execution path.

Latency is the round-trip time between your VPS and the broker server. Lower latency can help execution-sensitive EAs, but the best VPS is not only the fastest one. It is the one that stays online and keeps MT5 reporting.

What to test

TestWhat it shows
Ping to brokerApproximate network distance.
Order modification speedReal workflow latency under MT5.
Terminal uptimeWhether MT5 keeps running.
Reporter heartbeatWhether monitoring still sees the account.

Pick location by broker

Choose a VPS near the broker server, not near your house. If the broker server is in London, test London providers first. If it is in New York, test New York. The best location is empirical.

Reliability beats a tiny ping improvement

A 3 ms improvement does not matter if Windows updates restart the terminal or the VPS drops network. After latency, monitor stale terminals and account heartbeat. For a full setup guide, read Best VPS setup for MT5 EAs.

Measure more than one ping number

MT5 exposes the last known terminal-to-trade-server ping, which is useful but incomplete. End-to-end execution also includes EA processing, network variability, broker handling and the time until an order result returns. Record terminal ping, order round-trip time where safe, rejection codes and the spread or market condition at the time.

Measure from the VPS that actually runs the EA. A laptop speed test says little about the production path. Run samples during the sessions the strategy trades, because routing and load can vary by time. Prefer distributions—median, 95th percentile and worst sustained window—over one best screenshot.

A practical test plan

Collect terminal ping every minute for at least a representative trading week. Mark reconnects and VPS CPU or memory pressure. For a permitted demo or low-risk test, timestamp order submission and result receipt without turning the test into uncontrolled live exposure. Compare two candidate regions using the same broker server and the same method.

If region A has a 12 ms median and 90 ms 95th percentile while region B has a 20 ms median and 28 ms 95th percentile, the second may be operationally better for a latency-sensitive strategy: it is slower at the middle but much more consistent. The right choice depends on the strategy, not a universal “good ping” threshold.

Investigate the cause before migrating

  • Correlate latency spikes with terminal reconnects and host resource saturation.
  • Check whether every account on one VPS degrades at the same time.
  • Confirm the exact broker server; similar server names may route differently.
  • Re-test after any VPS, broker or network change.
  • Keep latency and stale-data alerts separate because they require different actions.

Continue with the VPS architecture guide and freshness monitoring guide. Latency measurement should lead to a reproducible decision, not a marketing claim based on the lowest observed number.

Control the test environment

Record VPS region, CPU plan, operating-system load, MT5 build, broker server and measurement time with every run. Pause unrelated maintenance tasks and note whether other terminals were active. If two providers are compared under different broker servers or sessions, the result cannot isolate the VPS.

Set an acceptance rule before looking at the numbers—for example, no disconnects, a bounded 95th percentile and enough resource headroom during the busiest session. Re-run after Windows updates, terminal changes or moving additional accounts onto the host. Keep a baseline so gradual degradation is visible. A one-off fast result is less valuable than a stable path that stays within the strategy’s operational budget.

Retain raw timestamped samples, not only a summary average. They let you align slow periods with CPU saturation, backups, terminal logs and broker events. When the path degrades, change one variable at a time—region, host plan or account density—then repeat the same collection window before declaring the problem solved.

Primary source

Technical reference: official MetaTrader documentation. Broker and prop-firm terms can differ; verify the rules for the exact account you operate.

Know when the VPS stops reporting.

xTriel watches MT5 heartbeat and account state from outside the terminal.

Connect MT5