Skip to main content

​
Overview

SLA (Service Level Agreement) tracking helps you monitor whether your services meet uptime and performance commitments. Pongo automatically calculates uptime percentages and displays them on dashboards alongside your defined SLA targets.

​
Configuring SLA Targets

Set SLA targets at the dashboard level: 
// pongo/dashboards/production.ts
import type { DashboardConfig } from "@/lib/config-types";

export default {
  name: "Production",
  slug: "production",
  public: true,
  slaTarget: 99.9,  // 99.9% uptime target
  monitors: ["api", "database", "cdn"],
} satisfies DashboardConfig;

​
Common SLA Targets

SLA TargetDowntime per YearDowntime per MonthDowntime per Day
99%3.65 days7.31 hours14.40 minutes
99.5%1.83 days3.65 hours7.20 minutes
99.9%8.77 hours43.83 minutes1.44 minutes
99.95%4.38 hours21.92 minutes43.20 seconds
99.99%52.60 minutes4.38 minutes8.64 seconds
99.999%5.26 minutes26.30 seconds0.86 seconds

​
Uptime Calculation

Pongo calculates uptime percentage using check results: 
const uptime = (successfulChecks / totalChecks) * 100;

​
Status Classification

  • Up: Counts toward successful checks
  • Down: Counts as failure
  • Degraded: Configurable (typically counts as partial success)

​
Time Windows

Uptime is calculated across multiple time windows:
  • Last 24 hours
  • Last 7 days
  • Last 30 days
  • Last 90 days
  • All time
This allows you to track both recent performance and long-term trends.

​
Response Time Tracking

Beyond uptime, SLA tracking includes response time metrics:

​
Latency Percentiles

  • P50 (median): 50% of requests are faster than this
  • P95: 95% of requests are faster than this (excludes outliers)
  • P99: 99% of requests are faster than this (catches slowest 1%)

​
Average Response Time

Calculated across all successful checks in the time window.

​
Status Distribution

Dashboards display how much time services spend in each state: 
Up: 99.2% (23h 48m)
Degraded: 0.5% (7m)
Down: 0.3% (4m)
This breakdown helps identify:
  • Chronic issues: High percentage of “down” time
  • Performance problems: Frequent “degraded” status
  • Stability: Consistent “up” status

​
Dashboard Display

​
SLA Indicator

Dashboards show current uptime against the target: 
Uptime: 99.87% / 99.9% target ✓
  • Green checkmark: Meeting SLA
  • Red warning: Below SLA target

​
Visual Components

  1. Uptime Badge: Prominent display of current uptime percentage
  2. SLA Progress Bar: Visual indicator of target achievement
  3. Status Distribution Pie Chart: Breakdown of up/degraded/down time
  4. Response Time Chart: Historical response time trends with P50/P95/P99 lines
  5. Uptime Bars: Color-coded timeline showing service status over time

​
Monitor-Level Metrics

Each monitor in a dashboard tracks its own metrics: 
{
  monitorId: "api",
  uptime: 99.95,
  avgResponseTime: 145,
  p50: 120,
  p95: 280,
  p99: 450,
  statusDistribution: {
    up: 99.95,
    degraded: 0.03,
    down: 0.02
  }
}
This allows you to:
  • Identify which services are impacting overall SLA
  • Prioritize improvements based on worst performers
  • Track individual service commitments

​
Setting Performance Thresholds

Use monitor handlers to define what constitutes “degraded” vs “up”: 
export default monitor({
  name: "API Latency SLA",
  interval: "1m",
  
  async handler() {
    const start = Date.now();
    const res = await fetch("https://api.example.com");
    const responseTime = Date.now() - start;
    
    return {
      status: !res.ok 
        ? "down" 
        : responseTime > 2000 
          ? "degraded"  // SLA: response time must be < 2s
          : "up",
      responseTime,
      statusCode: res.status,
    };
  },
});
This ensures your uptime percentage reflects both availability and performance requirements.

​
Alert Integration

Combine SLA tracking with alerts to notify when targets are at risk: 
export default monitor({
  name: "API",
  interval: "1m",
  
  alerts: [
    {
      id: "sla-at-risk",
      name: "SLA At Risk",
      condition: { consecutiveFailures: 5 },  // 5 minutes down = risk
      channels: ["slack"],
      severity: "warning",
    },
    {
      id: "sla-breach",
      name: "SLA Breach",
      condition: { consecutiveFailures: 15 },  // 15 minutes = breach
      channels: ["pagerduty"],
      severity: "critical",
    },
  ],
  
  async handler() { /* ... */ },
});

​
Public Status Page Display

When public: true, status pages show:
  1. Current uptime percentage with SLA target
  2. Response time charts with historical trends
  3. Latency percentiles (P50, P95, P99)
  4. Status distribution pie chart
  5. Uptime bars showing 90-day history
This transparency helps users understand service reliability and sets expectations.

​
Best Practices

Set Realistic Targets

Start with achievable SLA targets (99% or 99.5%) and increase as reliability improves.

Include Degraded States

Don’t just track up/down—use “degraded” for slow responses or partial outages.

Monitor Multiple Time Windows

Track 24h, 7d, and 30d uptime to spot both acute and chronic issues.

Alert Before Breach

Set up early warning alerts before you’re in danger of missing SLA commitments.

​
Calculating SLA Credits

Many SLA agreements include service credits for downtime. Use check results to calculate: 
// Example: Calculate monthly SLA compliance
const monthlyChecks = await getCheckResults({
  startDate: startOfMonth(),
  endDate: endOfMonth(),
});

const totalChecks = monthlyChecks.length;
const upChecks = monthlyChecks.filter(c => c.status === "up").length;
const uptime = (upChecks / totalChecks) * 100;

if (uptime < 99.9) {
  const breachMinutes = ((99.9 - uptime) / 100) * 43800; // minutes in month
  const creditPercent = calculateCredit(breachMinutes);
  console.log(`SLA breach: ${creditPercent}% credit owed`);
}

Dashboards

Configure dashboards with SLA targets

Monitors

Create monitors with performance thresholds

Alerts

Set up alerts for SLA breach risks

Status Pages

Display SLA metrics on public status pages

Build docs developers (and LLMs) love

Get started for freeTalk to us