telex-server-monitor-sdk/dist/services/periodicMonitorServices/cpuMonitor.js

SDK for Telex Server Monitor - A lightweight agent that collects and reports server CPU metrics to the Telex monitoring platform

import { getStoreData, logger } from "../../index.js";
import { CollectorService } from "../../metrics/collector.js";
import { connectToIntegrationServer, OutGoingMessageReplyType, sendReply, } from "../zeromqService.js";
export async function startCpuMonitoring() {
    const storeData = getStoreData();
    const channelId = storeData?.outputChannelId;
    if (!channelId) {
        logger.error("Channel ID not found. Cannot start CPU monitoring.");
        // Return a dummy interval that we'll clear immediately
        const dummyInterval = setInterval(() => { }, 1000);
        clearInterval(dummyInterval);
        return dummyInterval;
    }
    // connect to integration server
    await connectToIntegrationServer(channelId);
    logger.info("Starting CPU threshold monitoring");
    // Reset alert tracking
    let lastAlertTime = 0;
    // start periodic monitoring
    return setInterval(async () => {
        try {
            const metrics = await CollectorService.getMetrics();
            const storeData = getStoreData();
            const threshold = storeData?.cpuThreshold || 80;
            const currentTime = Date.now();
            // Only send alerts if CPU usage is above threshold
            if (metrics.cpu && metrics.cpu.usage > threshold) {
                // Determine alert severity
                const isCritical = metrics.cpu.usage > threshold + 10;
                // Only send alerts if it has been more than 5 minutes since the last alert of the same severity
                // This prevents alert flooding
                const alertKey = isCritical ? "critical" : "warning";
                const cooldownPeriod = 5 * 60 * 1000; // 5 minutes in ms
                if (!lastAlertTime || currentTime - lastAlertTime > cooldownPeriod) {
                    // Log and send the alert
                    if (isCritical) {
                        logger.error(`CRITICAL: CPU usage at ${metrics.cpu.usage.toFixed(1)}%, exceeding threshold of ${threshold}%`);
                    }
                    else {
                        logger.warn(`WARNING: CPU usage at ${metrics.cpu.usage.toFixed(1)}%, exceeding threshold of ${threshold}%`);
                    }
                    // Send alert via ZeroMQ
                    await sendCpuAlert(channelId, metrics, threshold, isCritical);
                    // Update last alert time
                    lastAlertTime = currentTime;
                }
                else {
                    logger.info(`Suppressing ${alertKey} CPU alert (${metrics.cpu.usage.toFixed(1)}%) - cooldown period active`);
                }
            }
        }
        catch (error) {
            logger.error(`Error in CPU monitoring: ${error.message}`);
        }
    }, 60000); // Check every minute
}
// Send a CPU threshold alert to the integration server
export async function sendCpuAlert(channelId, metrics, threshold, isCritical) {
    const severityEmoji = isCritical ? "🔥" : "⚠️";
    const severityText = isCritical ? "CRITICAL" : "WARNING";
    const alertMessage = {
        metrics,
        threshold,
        severity: isCritical ? "critical" : "warning",
        message: `${severityEmoji} ${severityText}: CPU Usage Alert ${severityEmoji}\n\nCPU usage (${metrics.cpu.usage.toFixed(1)}%) has exceeded the threshold (${threshold}%)\n\nServer: ${getStoreData()?.serverName || "Unknown"}\nCPU Cores: ${metrics.cpu?.cores || "N/A"}\nTimestamp: ${new Date().toLocaleString()}`,
    };
    await sendReply(channelId, alertMessage, OutGoingMessageReplyType.cpuThresholdAlertReply);
}