dora-metrics-and-devops-performance

name	DORA Metrics and DevOps Performance
description	The agent implements DORA metrics tracking for measuring and improving software delivery performance. Use when establishing engineering metrics, benchmarking teams, or driving DevOps transformation.
category	devops

DORA Metrics and DevOps Performance

Purpose

DORA (DevOps Research and Assessment) metrics are the industry standard for measuring software delivery performance. Google's research across thousands of organizations identified four key metrics that predict:

• Organizational performance (profitability, market share)
• Non-commercial performance (quality, customer satisfaction)
• Team well-being and reduced burnout

Elite performers who meet reliability targets are 2.3x more likely to use trunk-based development and continuous delivery practices.

Features

Metric	What It Measures	Elite Benchmark
Deployment Frequency	How often code reaches production	Multiple times per day
Lead Time for Changes	Time from commit to production	Less than 1 hour
Change Failure Rate	Percentage of deployments causing failures	0-15%
Time to Restore Service	Recovery time from incidents	Less than 1 hour

The Four Key Metrics

1. Deployment Frequency

Definition: How often your organization deploys code to production.

// Deployment frequency calculation
interface DeploymentData {
  timestamp: Date;
  environment: string;
  service: string;
  success: boolean;
}

function calculateDeploymentFrequency(
  deployments: DeploymentData[],
  periodDays: number = 30
): { frequency: string; deploymentsPerDay: number } {
  const productionDeployments = deployments.filter(
    d => d.environment === 'production' && d.success
  );

  const deploymentsPerDay = productionDeployments.length / periodDays;

  let frequency: string;
  if (deploymentsPerDay >= 1) {
    frequency = 'elite'; // Multiple times per day or daily
  } else if (deploymentsPerDay >= 1/7) {
    frequency = 'high'; // Weekly to daily
  } else if (deploymentsPerDay >= 1/30) {
    frequency = 'medium'; // Monthly to weekly
  }  {
    frequency = ; 
  }

   { frequency, deploymentsPerDay };
}

2. Lead Time for Changes

Definition: Time from code commit to code running in production.

// Lead time calculation
interface ChangeData {
  commitTimestamp: Date;
  deployTimestamp: Date;
  commitSha: string;
  prNumber?: number;
}

function calculateLeadTime(changes: ChangeData[]): {
  medianHours: number;
  p90Hours: number;
  performance: string;
} {
  const leadTimes = changes.map(c =>
    (c.deployTimestamp.getTime() - c.commitTimestamp.getTime()) / (1000 * 60 * 60)
  );

  leadTimes.sort((a, b) => a - b);

  const median = leadTimes[Math.floor(leadTimes.length / 2)];
  const p90 = leadTimes[Math.floor(leadTimes.length * 0.9)];

  let performance: string;
  if (median < 1) {
    performance = 'elite'; 
  }   (median < ) {
    performance = ; 
  }   (median < ) {
    performance = ; 
  }  {
    performance = ; 
  }

   { : median, : p90, performance };
}

3. Change Failure Rate

Definition: Percentage of deployments that result in degraded service requiring remediation.

// Change failure rate calculation
interface DeploymentOutcome {
  deploymentId: string;
  timestamp: Date;
  success: boolean;
  causedIncident: boolean;
  requiredRollback: boolean;
  requiredHotfix: boolean;
}

function calculateChangeFailureRate(deployments: DeploymentOutcome[]): {
  rate: number;
  performance: string;
} {
  const total = deployments.length;
  const failures = deployments.filter(d =>
    d.causedIncident || d.requiredRollback || d.requiredHotfix
  ).length;

  const rate = (failures / total) * 100;

  let performance: string;
  if (rate <= 15) {
    performance = 'elite'; // 0-15%
  } else if (rate <= 30) {
    performance = 'high'; // 16-30%
  } else if (rate <= 45) {
    performance = ; 
  }  {
    performance = ; 
  }

   { rate, performance };
}

4. Time to Restore Service (MTTR)

Definition: How long it takes to restore service when an incident occurs.

// MTTR calculation
interface Incident {
  id: string;
  startTime: Date;
  resolvedTime: Date;
  severity: 'critical' | 'major' | 'minor';
  service: string;
}

function calculateMTTR(incidents: Incident[]): {
  medianHours: number;
  performance: string;
  byService: Record<string, number>;
} {
  const restorationTimes = incidents.map(i =>
    (i.resolvedTime.getTime() - i.startTime.getTime()) / (1000 * 60 * 60)
  );

  restorationTimes.sort((a, b) => a - b);
  const median = restorationTimes[Math.floor(restorationTimes.length / 2)];

  let performance: string;
  if (median < 1) {
    performance = 'elite'; 
  }   (median < ) {
    performance = ; 
  }   (median < ) {
    performance = ; 
  }  {
    performance = ; 
  }

  
   : <, []> = {};
   ( incident  incidents) {
     (!byService[incident.]) byService[incident.] = [];
     hours = (incident..() - incident..()) / ( *  * );
    byService[incident.].(hours);
  }

   : <, > = {};
   ( [service, times]  .(byService)) {
    times.( a - b);
    serviceMedians[service] = times[.(times. / )];
  }

   { : median, performance, : serviceMedians };
}

Performance Levels (2024 Benchmarks)

Level	Deploy Freq	Lead Time	Change Failure	MTTR
Elite	Multiple/day	< 1 hour	0-15%	< 1 hour
High	Daily-Weekly	1 day - 1 week	16-30%	< 1 day
Medium	Weekly-Monthly	1 week - 1 month	16-30%	< 1 day
Low	Monthly+	1-6 months	16-30%	< 1 week

Key Insight (2024 DORA Report): Elite performers are 2.3x more likely to meet reliability targets when using trunk-based development.

Measurement Implementation

GitHub Actions DORA Workflow

# .github/workflows/dora-metrics.yml
name: DORA Metrics Collection

on:
  schedule:
    - cron: '0 0 * * 0'  # Weekly on Sunday
  workflow_dispatch:

jobs:
  collect-metrics:
    runs-on: ubuntu-latest
    steps:
      - uses: actions/checkout@v4

      - name: Collect Deployment Data
        id: deployments
        uses: actions/github-script@v7
        with:
          script: |
            const thirtyDaysAgo = new Date();
            thirtyDaysAgo.setDate(thirtyDaysAgo.getDate() - 30);

            // Get workflow runs (deployments)
            const { data: runs } = await github.rest.actions.listWorkflowRuns({
              owner: context.repo.owner,
              repo: context.repo.repo,
              workflow_id: 'deploy.yml',
              created: `>=${thirtyDaysAgo.toISOString()}`,
               
            

                
                
            

               
                 

             {
               ,
               ,
                    
                              
                               
            }

           
         
         
        
           
               
             {   }   
               
               
               
               
               
               
            

                
               
                 
            

                 {
                  
                  
                         
            }

                 
                   

             {
               ,
                    
                               
                                
            }

         
         
          

                
          
           [    ]    
           [    ]    
          

         
         
        
           
           

Custom Metrics Collection Script

// scripts/collect-dora-metrics.ts
import { Octokit } from '@octokit/rest';

interface DORAMetrics {
  period: { start: Date; end: Date };
  deploymentFrequency: {
    count: number;
    perDay: number;
    performance: 'elite' | 'high' | 'medium' | 'low';
  };
  leadTime: {
    medianHours: number;
    p90Hours: number;
    performance: 'elite' | 'high' | 'medium' | 'low';
  };
  changeFailureRate: {
    total: number;
    failures: number;
    rate: number;
    performance: 'elite' | 'high' | 'medium' | 'low';
  };
  mttr: {
    medianHours: number;
    incidentCount: number;
    performance: 'elite' | 'high' | 'medium' | 'low';
  };
  :  |  |  | ;
}

  {
   : ;
   : ;
   : ;

  () {
    . =  ({ : token });
    . = owner;
    . = repo;
  }

   (:  = ): <> {
     end =  ();
     start =  ();
    start.(start.() - periodDays);

     [deployments, prs, incidents] =  .([
      .(start, end),
      .(start, end),
      .(start, end)
    ]);

    
     deploymentFrequency = .(deployments, periodDays);
     leadTime = .(prs);
     changeFailureRate = .(deployments, incidents);
     mttr = .(incidents);

    
     performances = [
      deploymentFrequency.,
      leadTime.,
      changeFailureRate.,
      mttr.
    ];

     overallPerformance = .(performances);

     {
      : { start, end },
      deploymentFrequency,
      leadTime,
      changeFailureRate,
      mttr,
      overallPerformance
    };
  }

    () {
     { data } =  ...({
      : .,
      : .,
      : ,
      : 
    });
     data.;
  }

    () {
     { data } =  ...({
      : .,
      : .,
      : ,
      : ,
      : 
    });
     data.(
      pr. &&
       (pr.) >= start &&
       (pr.) <= end
    );
  }

    () {
    
    
     { data } =  ...({
      : .,
      : .,
      : ,
      : ,
      : start.()
    });
     data;
  }

   () {
     successful = deployments.( d. === );
     perDay = successful. / periodDays;

     {
      : successful.,
      perDay,
      : .(perDay)
    };
  }

   () {
     times = prs.( {
       created =  (pr.);
       merged =  (pr.);
       (merged.() - created.()) / ( *  * );
    });

    times.( a - b);
     median = times[.(times. / )] || ;
     p90 = times[.(times. * )] || ;

     {
      : median,
      : p90,
      : .(median)
    };
  }

   () {
     total = deployments.( d. === ).;
     failures = incidents.; 

     rate = total >  ? (failures / total) *  : ;

     {
      total,
      failures,
      rate,
      : .(rate)
    };
  }

   () {
     times = incidents
      .( i.)
      .( {
         opened =  (i.);
         closed =  (i.);
         (closed.() - opened.()) / ( *  * );
      });

    times.( a - b);
     median = times[.(times. / )] || ;

     {
      : median,
      : incidents.,
      : .(median)
    };
  }

   (: ):  |  |  |  {
     (perDay >= )  ;
     (perDay >= /)  ;
     (perDay >= /)  ;
     ;
  }

   (: ):  |  |  |  {
     (hours < )  ;
     (hours < )  ;
     (hours < )  ;
     ;
  }

   (: ):  |  |  |  {
     (rate <= )  ;
     (rate <= )  ;
     (rate <= )  ;
     ;
  }

   (: ):  |  |  |  {
     (hours < )  ;
     (hours < )  ;
     (hours < )  ;
     ;
  }

   (: []):  |  |  |  {
     scores = { : , : , : , :  };
     avg = performances.( sum + scores[p  keyof  scores], ) / performances.;

     (avg >= )  ;
     (avg >= )  ;
     (avg >= )  ;
     ;
  }
}


 collector =  (
  process..!,
  ,
  
);

 metrics =  collector.();
.(.(metrics, , ));

Grafana Dashboard Configuration

{
  "dashboard": {
    "title": "DORA Metrics Dashboard",
    "panels": [
      {
        "title": "Deployment Frequency",
        "type": "stat",
        "targets": [
          {
            "expr": "sum(increase(deployments_total{environment=\"production\"}[30d])) / 30",
            "legendFormat": "Deploys/day"
          }
        ],
        "fieldConfig": {
          "defaults": {
            "thresholds": {
              "steps": [
                { "value": 0, "color": "red" }
                     
                     
                     
              
            
          
        
      
      
         
         
         
          
             
             
          
        
         
           
             
             
               
                     
                     
                     
                     
              
            
          
        
      
      
         
         
         
          
             
             
          
        
         
           
             
             
             
             
               
                     
                     
                     
                     
              
            
          
        
      
      
         
         
         
          
             
             
          
        
         
           
             
             
               
                     
                     
                     
                     
              
            
          
        
      
    
  

Tools and Platforms

Tool	Type	Features
Four Keys (Google)	Open Source	GitHub/GitLab integration, BigQuery
LinearB	Commercial	Git analytics, workflow metrics
Sleuth	Commercial	Deploy tracking, change intelligence
Faros AI	Commercial	Multi-source aggregation
Propelo	Commercial	SDLC insights
Jellyfish	Commercial	Engineering management

Four Keys Setup (Google)

# Deploy Four Keys to GCP
git clone https://github.com/dora-team/fourkeys.git
cd fourkeys

# Configure
export PROJECT_ID="my-project"
export REGION="us-central1"

# Deploy
./setup/setup.sh

# Configure webhook for GitHub events
# Add to GitHub repo settings: https://<REGION>-<PROJECT_ID>.cloudfunctions.net/github-parser

Improvement Strategies

Improving Deployment Frequency

Current	Target	Strategy
Monthly	Weekly	Automate deployments, reduce batch size
Weekly	Daily	Feature flags, trunk-based development
Daily	Multiple/day	Continuous deployment, small PRs

Improving Lead Time

Bottleneck	Solution
Long code reviews	Smaller PRs, async reviews, automation
Manual testing	Automated tests, shift-left
Manual deployments	CI/CD automation
Environment issues	Infrastructure as code

Reducing Change Failure Rate

Problem	Solution
Insufficient testing	Increase coverage, add integration tests
Big bang releases	Feature flags, canary releases
Lack of review	Automated checks, required reviews
Poor monitoring	Better observability, alerting

Reducing MTTR

Improvement	Impact
Runbooks	Faster diagnosis
Feature flags	Instant rollback
Observability	Faster root cause
Chaos engineering	Proactive resilience

Best Practices

1. Measure Consistently

// Standardized metric definitions
const METRIC_DEFINITIONS = {
  deploymentFrequency: {
    source: 'GitHub Actions',
    filter: 'workflow=deploy.yml, conclusion=success',
    aggregation: 'count per day'
  },
  leadTime: {
    source: 'GitHub PRs',
    measurement: 'created_at to merged_at',
    aggregation: 'median'
  },
  changeFailureRate: {
    source: 'GitHub Issues + Deployments',
    filter: 'label=incident, within 24h of deployment',
    aggregation: 'incidents / deployments * 100'
  },
  mttr: {
    source: 'PagerDuty',
    measurement: 'triggered_at to resolved_at',
    aggregation: 'median'
  }
};

2. Set Realistic Goals

# Quarterly improvement targets
q1_2024:
  deployment_frequency:
    current: 0.5/day
    target: 1.0/day
    improvement: 100%
  lead_time:
    current: 48h
    target: 24h
    improvement: 50%
  change_failure_rate:
    current: 25%
    target: 20%
    improvement: 20%
  mttr:
    current: 4h
    target: 2h
    improvement: 50%

3. Avoid Gaming Metrics

Gaming Behavior	Why It's Bad	Better Approach
Deploying empty commits	Fake frequency	Track meaningful changes
Not labeling incidents	Hide failures	Blameless culture
Splitting PRs artificially	Fake lead time	Focus on value
Rushing fixes	Lower quality	Fix root cause

Use Cases

1. Team Performance Review

// Quarterly DORA review
async function quarterlyReview(team: string) {
  const metrics = await collectMetrics({ team, period: '90d' });

  return {
    summary: {
      overallPerformance: metrics.overallPerformance,
      strongestMetric: findStrongest(metrics),
      improvementArea: findWeakest(metrics)
    },
    comparison: {
      vsLastQuarter: await compareToLastQuarter(team, metrics),
      vsIndustry: compareToIndustryBenchmarks(metrics)
    },
    recommendations: generateRecommendations(metrics)
  };
}

2. DevOps Transformation Tracking

// Track transformation progress
const transformationGoals = {
  phase1: { // Foundation
    deploymentFrequency: 'weekly',
    leadTime: '< 1 week'
  },
  phase2: { // Acceleration
    deploymentFrequency: 'daily',
    leadTime: '< 1 day',
    changeFailureRate: '< 30%'
  },
  phase3: { // Excellence
    deploymentFrequency: 'multiple/day',
    leadTime: '< 1 hour',
    changeFailureRate: '< 15%',
    mttr: '< 1 hour'
  }
};

Related Skills

• devops/github-actions - CI/CD automation
• devops/observability - Monitoring and metrics
• testing/comprehensive-testing - Quality gates
• devops/feature-flags - Progressive delivery

---

Think Omega. Build Omega. Be Omega.