Great question—dealing with inherent baseline packet loss while trying to spot actual network outages is such a common headache, especially when the built-in ping.exe feels too limited in how you can tweak and analyze results. Let’s walk through practical alternatives and custom setups that’ll give you far more control over your ICMP monitoring:

This is my go-to for scenarios like yours because you can tailor every part of the check to your exact needs:

Python (Cross-Platform) #

Libraries like ping3 or scapy let you build a custom monitor from scratch. For example, you could:

• Spin up two parallel check threads (matching your current setup)
• Set a 950ms timeout per check
• Add logic to only flag an outage if both threads fail to get a response for 2+ consecutive attempts (since the odds of your 1.7% baseline causing both to drop at the same time are tiny—~0.03%—this eliminates false positives)
• Log results or update your status indicator in real-time

Here’s a quick snippet to illustrate the core idea:

import ping3
import threading
import time

def ping_check():
    response_time = ping3.ping('8.8.8.8', timeout=0.95)
    return response_time is not None

def monitor():
    outage_count = 0
    while True:
        # Run two parallel checks
        thread1 = threading.Thread(target=ping_check)
        thread2 = threading.Thread(target=ping_check)
        thread1.start()
        thread2.start()
        thread1.join()
        thread2.join()
        # Fetch results (note: adjust if using a thread result wrapper)
        result1 = ping_check()
        result2 = ping_check()

        if not result1 and not result2:
            outage_count += 1
            if outage_count >= 2:
                print("*NETWORK OUTAGE DETECTED*")
        else:
            outage_count = 0
            print("Network stable")
        time.sleep(1)

if __name__ == "__main__":
    monitor()

PowerShell (Windows-Native) #

If you prefer sticking to Windows tools, Test-Connection is way more flexible than ping.exe. You can script similar logic to your two-process setup, with built-in support for timeouts and parallel checks:

$target = "8.8.8.8"
$timeoutMs = 950
$outageThreshold = 2
$consecutiveFailures = 0

while ($true) {
    # Run two parallel ICMP checks
    $result1 = Test-Connection -ComputerName $target -Count 1 -TimeoutSeconds ($timeoutMs/1000) -ErrorAction SilentlyContinue
    $result2 = Test-Connection -ComputerName $target -Count 1 -TimeoutSeconds ($timeoutMs/1000) -ErrorAction SilentlyContinue

    if (-not $result1 -and -not $result2) {
        $consecutiveFailures++
        if ($consecutiveFailures -ge $outageThreshold) {
            Write-Host "⚠️ NETWORK OUTAGE DETECTED ⚠️" -ForegroundColor Red
        }
    } else {
        $consecutiveFailures = 0
        Write-Host "✅ Network stable" -ForegroundColor Green
    }
    Start-Sleep -Seconds 1
}

If scripting isn’t your thing, there are lightweight tools designed for this exact use case:

• PingInfoView (Windows): Lets you monitor multiple targets simultaneously, set custom timeouts, and configure alerts based on consecutive packet loss (instead of single drops). You can even export logs to analyze your baseline loss over time.
• Nagios Core (Cross-Platform): While it’s more robust, you can set up a simple ICMP check with custom thresholds—configure it to only alert if both of your parallel checks fail consecutively, matching your current logic.
• PRTG Network Monitor: Has a free tier for personal use, with built-in ICMP sensors that let you set warning/critical thresholds based on packet loss percentage and consecutive failures.

If you want to stick with ping.exe but gain better control, wrap it in a simple batch script to parse output and track consecutive failures:

@echo off
setlocal enabledelayedexpansion

set "target=8.8.8.8"
set "timeout=950"
set "outageThreshold=2"
set "consecutiveFails=0"

:loop
:: Run two ping processes in parallel and capture results
start /b ping -n 1 -w %timeout% %target% > ping1.txt
start /b ping -n 1 -w %timeout% %target% > ping2.txt
:: Wait for both to finish
timeout /t 1 /nobreak > nul

:: Check results
set "result1=fail"
set "result2=fail"
findstr /i "Reply from" ping1.txt > nul && set "result1=pass"
findstr /i "Reply from" ping2.txt > nul && set "result2=pass"

if "!result1!"=="fail" if "!result2!"=="fail" (
    set /a consecutiveFails+=1
    if !consecutiveFails! geq %outageThreshold% (
        echo *NETWORK OUTAGE DETECTED*
    )
) else (
    set consecutiveFails=0
    echo Network stable
)

del ping1.txt ping2.txt
goto loop

All these approaches let you move beyond the limited control of raw ping.exe and accurately distinguish between your baseline 1.7% packet loss and actual network interruptions. Pick the one that fits your comfort level with coding or pre-built tools!

内容的提问来源于stack exchange，提问作者Llamageddon