Latency and percentiles
P95
The 95th percentile. In simple terms, 95% of measured values were this number or lower, and the slowest 5% were higher. It is useful because it shows “bad but common enough” delays without being distorted by one extreme outlier.
Average / Avg
The usual mean value. It is easy to understand, but it can hide short periods of poor performance or be distorted by very large spikes.
Min / Max
The best and worst observed values. Max is helpful for spotting extreme events, but one maximum value alone does not prove the connection was always bad.
Ping / Background latency
Approximate network response time between the terminal and broker access point. Lower is generally better, especially for active trading and scalping.
Jitter
How much latency changes over time. Stable 40 ms can be better than a connection jumping between 20 ms and 400 ms.
Trade execution metrics
Execution time, ms
How long the terminal reports it took to complete or reject a trade request. High values can be caused by local network, VPS route, broker infrastructure, market conditions, or platform load.
Average execution
The typical execution time across detected trades. Use it as a baseline, then review spikes and rejected trades for the real pain points.
Slowest / Max execution
The worst detected execution delay. Check its timestamp against disconnects, NTP drift, broker rejections, or VPS issues.
Spikes above P95
Trades that were slower than the normal upper range. These are useful candidates for investigation because they represent unusually slow execution.
Connectivity metrics
Connection drops / Disconnects
Times when the terminal lost broker connectivity. Frequent drops can affect market data, order sending, EA behavior, and account synchronization.
Reconnects
Times when the terminal restored the connection. Disconnects without reconnects may indicate a longer outage or an incomplete log window.
Reconnect coverage
The share of disconnect events that were followed by a detected reconnect. Higher is better; lower coverage means outages may not have recovered cleanly inside the log.
Access point / Route
The broker server or network route used by the terminal. Different access points can have different latency and reliability.
Cellular share
How much mobile activity appears to use cellular connectivity. Cellular networks can be reliable, but they are usually less predictable than a stable VPS or wired connection.
Reliability and terminal health
Crash ratio
The share of terminal exits that look abnormal. A higher ratio can suggest platform instability, bad drivers, resource pressure, or problematic add-ons.
System warnings
Terminal or environment warnings detected in the log. Review them when they happen near trade delays or disconnects.
Subsystem failures
Problems reported by components such as Market, Virtual Hosting, LiveUpdate, or network subsystems. These may explain missing services or update/connectivity issues.
EA failures
Detected Expert Advisor load/runtime failures. If trade issues coincide with EA failures, check EA settings, dependencies, permissions, and symbol/timeframe configuration.
Trade result metrics
Rejected / Failed
Trade requests that were not completed successfully. The reason matters: timeout, off quotes, market closed, no money, invalid volume, and requote mean different things.
Rejection reason
The broker or terminal explanation for why an order failed. Use this together with timestamp, symbol, spread, connectivity, and execution delay.
Pending orders placed
Limit or stop orders that were accepted by the broker. These are different from market orders because execution happens later when price conditions are met.
Infrastructure diagnosis
Infra risk score
A practical summary score based on detected infrastructure signals. It is a triage indicator, not a final verdict.
NTP drift
Difference between local time and reference time. Large drift can make event correlation difficult and may indicate time-sync issues on a PC or VPS.
NTP anomalies
Large time-sync deviations, usually above one second. If these appear often, check OS time synchronization and VPS clock settings.
Tracked environments
Distinct terminal/device environments seen in the log. Multiple environments can be normal, but they can also explain different IPs, routes, builds, or behavior.