DeWitt Gibson commited on
Commit
b7b516d
·
1 Parent(s): 04ef1ef

updating docs

Browse files
Files changed (2) hide show
  1. README.md +121 -0
  2. docs/README.md +555 -1
README.md CHANGED
@@ -301,6 +301,127 @@ docker run -p 8000:8000 \
301
 
302
  See [docker/README.md](docker/README.md) for complete Docker documentation.
303
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
304
  ## ⚙️ Configuration
305
 
306
  ### Environment Variables
 
301
 
302
  See [docker/README.md](docker/README.md) for complete Docker documentation.
303
 
304
+ ## ☁️ Cloud Deployment
305
+
306
+ LLMGuardian can be deployed on all major cloud platforms. Below are quick start guides for each provider. For detailed step-by-step instructions, see [PROJECT.md - Cloud Deployment Guides](PROJECT.md#cloud-deployment-guides).
307
+
308
+ ### AWS Deployment
309
+
310
+ **Option 1: ECS with Fargate (Recommended)**
311
+ ```bash
312
+ # Push to ECR and deploy
313
+ aws ecr get-login-password --region us-east-1 | docker login --username AWS --password-stdin YOUR_ACCOUNT_ID.dkr.ecr.us-east-1.amazonaws.com
314
+ aws ecr create-repository --repository-name llmguardian
315
+ docker tag llmguardian:latest YOUR_ACCOUNT_ID.dkr.ecr.us-east-1.amazonaws.com/llmguardian:latest
316
+ docker push YOUR_ACCOUNT_ID.dkr.ecr.us-east-1.amazonaws.com/llmguardian:latest
317
+ ```
318
+
319
+ **Other AWS Options:**
320
+ - AWS Lambda with Docker containers
321
+ - Elastic Beanstalk for PaaS deployment
322
+ - EKS for Kubernetes orchestration
323
+
324
+ ### Google Cloud Platform
325
+
326
+ **Cloud Run (Recommended)**
327
+ ```bash
328
+ # Build and deploy to Cloud Run
329
+ gcloud auth configure-docker
330
+ docker tag llmguardian:latest gcr.io/YOUR_PROJECT_ID/llmguardian:latest
331
+ docker push gcr.io/YOUR_PROJECT_ID/llmguardian:latest
332
+
333
+ gcloud run deploy llmguardian \
334
+ --image gcr.io/YOUR_PROJECT_ID/llmguardian:latest \
335
+ --platform managed \
336
+ --region us-central1 \
337
+ --allow-unauthenticated \
338
+ --memory 2Gi \
339
+ --port 8000
340
+ ```
341
+
342
+ **Other GCP Options:**
343
+ - Google Kubernetes Engine (GKE)
344
+ - App Engine for PaaS deployment
345
+
346
+ ### Microsoft Azure
347
+
348
+ **Azure Container Instances**
349
+ ```bash
350
+ # Create resource group and deploy
351
+ az group create --name llmguardian-rg --location eastus
352
+ az acr create --resource-group llmguardian-rg --name llmguardianacr --sku Basic
353
+ az acr login --name llmguardianacr
354
+
355
+ docker tag llmguardian:latest llmguardianacr.azurecr.io/llmguardian:latest
356
+ docker push llmguardianacr.azurecr.io/llmguardian:latest
357
+
358
+ az container create \
359
+ --resource-group llmguardian-rg \
360
+ --name llmguardian-container \
361
+ --image llmguardianacr.azurecr.io/llmguardian:latest \
362
+ --cpu 2 --memory 4 --ports 8000
363
+ ```
364
+
365
+ **Other Azure Options:**
366
+ - Azure App Service (Web App for Containers)
367
+ - Azure Kubernetes Service (AKS)
368
+ - Azure Functions
369
+
370
+ ### Vercel
371
+
372
+ **Serverless Deployment**
373
+ ```bash
374
+ # Install Vercel CLI and deploy
375
+ npm i -g vercel
376
+ vercel login
377
+ vercel --prod
378
+ ```
379
+
380
+ Create `vercel.json`:
381
+ ```json
382
+ {
383
+ "version": 2,
384
+ "builds": [{"src": "src/llmguardian/api/app.py", "use": "@vercel/python"}],
385
+ "routes": [{"src": "/(.*)", "dest": "src/llmguardian/api/app.py"}]
386
+ }
387
+ ```
388
+
389
+ ### DigitalOcean
390
+
391
+ **App Platform (Easiest)**
392
+ ```bash
393
+ # Using doctl CLI
394
+ doctl auth init
395
+ doctl apps create --spec .do/app.yaml
396
+ ```
397
+
398
+ **Other DigitalOcean Options:**
399
+ - DigitalOcean Kubernetes (DOKS)
400
+ - Droplets with Docker
401
+
402
+ ### Platform Comparison
403
+
404
+ | Platform | Best For | Ease of Setup | Estimated Cost |
405
+ |----------|----------|---------------|----------------|
406
+ | **GCP Cloud Run** | Startups, Auto-scaling | ⭐⭐⭐⭐⭐ Easy | $30-150/mo |
407
+ | **AWS ECS** | Enterprise, Flexibility | ⭐⭐⭐ Medium | $50-200/mo |
408
+ | **Azure ACI** | Microsoft Ecosystem | ⭐⭐⭐⭐ Easy | $50-200/mo |
409
+ | **Vercel** | API Routes, Serverless | ⭐⭐⭐⭐⭐ Very Easy | $20-100/mo |
410
+ | **DigitalOcean** | Simple, Predictable | ⭐⭐⭐⭐ Easy | $24-120/mo |
411
+
412
+ ### Prerequisites for Cloud Deployment
413
+
414
+ Before deploying to any cloud:
415
+
416
+ 1. **Prepare Environment Variables**: Copy `.env.example` to `.env` and configure
417
+ 2. **Build Docker Image**: `docker build -t llmguardian:latest -f docker/dockerfile .`
418
+ 3. **Set Up Cloud CLI**: Install and authenticate with your chosen provider
419
+ 4. **Configure Secrets**: Use cloud secret managers (AWS Secrets Manager, Azure Key Vault, GCP Secret Manager)
420
+ 5. **Enable HTTPS**: Configure SSL/TLS certificates
421
+ 6. **Set Up Monitoring**: Enable cloud-native monitoring and logging
422
+
423
+ For complete deployment guides with step-by-step instructions, configuration examples, and best practices, see **[PROJECT.md - Cloud Deployment Guides](PROJECT.md#cloud-deployment-guides)**.
424
+
425
  ## ⚙️ Configuration
426
 
427
  ### Environment Variables
docs/README.md CHANGED
@@ -1605,4 +1605,558 @@ response = requests.post(
1605
  ## API Status
1606
  Check status at: https://status.llmguardian.com # replace llmguardian.com with your domain
1607
 
1608
- Rate limits and API metrics available in dashboard.
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1605
  ## API Status
1606
  Check status at: https://status.llmguardian.com # replace llmguardian.com with your domain
1607
 
1608
+ Rate limits and API metrics available in dashboard.
1609
+
1610
+ ---
1611
+
1612
+ ## ☁️ Cloud Deployment Guides
1613
+
1614
+ LLMGuardian can be deployed on all major cloud platforms. This section provides comprehensive deployment guides for AWS, Google Cloud, Azure, Vercel, and DigitalOcean.
1615
+
1616
+ > **📘 For complete step-by-step instructions with all configuration details, see [PROJECT.md - Cloud Deployment Guides](../PROJECT.md#cloud-deployment-guides)**
1617
+
1618
+ ### Quick Start by Platform
1619
+
1620
+ #### AWS Deployment
1621
+
1622
+ **Recommended: ECS with Fargate**
1623
+
1624
+ ```bash
1625
+ # Push to ECR
1626
+ aws ecr get-login-password --region us-east-1 | docker login --username AWS --password-stdin YOUR_ACCOUNT_ID.dkr.ecr.us-east-1.amazonaws.com
1627
+ aws ecr create-repository --repository-name llmguardian --region us-east-1
1628
+
1629
+ # Tag and push
1630
+ docker tag llmguardian:latest YOUR_ACCOUNT_ID.dkr.ecr.us-east-1.amazonaws.com/llmguardian:latest
1631
+ docker push YOUR_ACCOUNT_ID.dkr.ecr.us-east-1.amazonaws.com/llmguardian:latest
1632
+
1633
+ # Create ECS cluster and deploy
1634
+ aws ecs create-cluster --cluster-name llmguardian-cluster --region us-east-1
1635
+ aws ecs register-task-definition --cli-input-json file://task-definition.json
1636
+ aws ecs create-service --cluster llmguardian-cluster --service-name llmguardian-service --task-definition llmguardian --desired-count 2
1637
+ ```
1638
+
1639
+ **Other AWS Options:**
1640
+ - **Lambda**: Serverless function deployment with Docker containers
1641
+ - **Elastic Beanstalk**: PaaS deployment with auto-scaling
1642
+ - **EKS**: Kubernetes orchestration for large-scale deployments
1643
+
1644
+ **Key Features:**
1645
+ - Auto-scaling with CloudWatch metrics
1646
+ - Load balancing with ALB/NLB
1647
+ - Secrets management with Secrets Manager
1648
+ - CloudWatch logging and monitoring
1649
+
1650
+ #### Google Cloud Platform
1651
+
1652
+ **Recommended: Cloud Run**
1653
+
1654
+ ```bash
1655
+ # Configure Docker for GCP
1656
+ gcloud auth configure-docker
1657
+
1658
+ # Build and push to GCR
1659
+ docker tag llmguardian:latest gcr.io/YOUR_PROJECT_ID/llmguardian:latest
1660
+ docker push gcr.io/YOUR_PROJECT_ID/llmguardian:latest
1661
+
1662
+ # Deploy to Cloud Run
1663
+ gcloud run deploy llmguardian \
1664
+ --image gcr.io/YOUR_PROJECT_ID/llmguardian:latest \
1665
+ --platform managed \
1666
+ --region us-central1 \
1667
+ --allow-unauthenticated \
1668
+ --memory 2Gi \
1669
+ --cpu 2 \
1670
+ --port 8000 \
1671
+ --min-instances 1 \
1672
+ --max-instances 10
1673
+ ```
1674
+
1675
+ **Other GCP Options:**
1676
+ - **GKE (Google Kubernetes Engine)**: Full Kubernetes control
1677
+ - **App Engine**: PaaS with automatic scaling
1678
+ - **Cloud Functions**: Event-driven serverless
1679
+
1680
+ **Key Features:**
1681
+ - Automatic HTTPS and custom domains
1682
+ - Built-in auto-scaling
1683
+ - Secret Manager integration
1684
+ - Cloud Logging and Monitoring
1685
+
1686
+ #### Microsoft Azure
1687
+
1688
+ **Recommended: Container Instances**
1689
+
1690
+ ```bash
1691
+ # Create resource group and registry
1692
+ az group create --name llmguardian-rg --location eastus
1693
+ az acr create --resource-group llmguardian-rg --name llmguardianacr --sku Basic
1694
+ az acr login --name llmguardianacr
1695
+
1696
+ # Push image
1697
+ docker tag llmguardian:latest llmguardianacr.azurecr.io/llmguardian:latest
1698
+ docker push llmguardianacr.azurecr.io/llmguardian:latest
1699
+
1700
+ # Deploy container instance
1701
+ az container create \
1702
+ --resource-group llmguardian-rg \
1703
+ --name llmguardian-container \
1704
+ --image llmguardianacr.azurecr.io/llmguardian:latest \
1705
+ --cpu 2 \
1706
+ --memory 4 \
1707
+ --dns-name-label llmguardian \
1708
+ --ports 8000 \
1709
+ --environment-variables LOG_LEVEL=INFO
1710
+ ```
1711
+
1712
+ **Other Azure Options:**
1713
+ - **App Service**: Web App for Containers with built-in CI/CD
1714
+ - **AKS (Azure Kubernetes Service)**: Managed Kubernetes
1715
+ - **Azure Functions**: Serverless with Python support
1716
+
1717
+ **Key Features:**
1718
+ - Azure Key Vault for secrets
1719
+ - Application Insights monitoring
1720
+ - Azure CDN integration
1721
+ - Auto-scaling capabilities
1722
+
1723
+ #### Vercel Deployment
1724
+
1725
+ **Serverless API Deployment**
1726
+
1727
+ ```bash
1728
+ # Install Vercel CLI
1729
+ npm i -g vercel
1730
+
1731
+ # Login and deploy
1732
+ vercel login
1733
+ vercel --prod
1734
+ ```
1735
+
1736
+ **Configuration** (`vercel.json`):
1737
+ ```json
1738
+ {
1739
+ "version": 2,
1740
+ "builds": [
1741
+ {
1742
+ "src": "src/llmguardian/api/app.py",
1743
+ "use": "@vercel/python"
1744
+ }
1745
+ ],
1746
+ "routes": [
1747
+ {
1748
+ "src": "/(.*)",
1749
+ "dest": "src/llmguardian/api/app.py"
1750
+ }
1751
+ ],
1752
+ "env": {
1753
+ "LOG_LEVEL": "INFO",
1754
+ "ENVIRONMENT": "production"
1755
+ }
1756
+ }
1757
+ ```
1758
+
1759
+ **Key Features:**
1760
+ - Automatic HTTPS and custom domains
1761
+ - Edge network deployment
1762
+ - Environment variable management
1763
+ - GitHub integration for auto-deploy
1764
+
1765
+ **Limitations:**
1766
+ - 10s execution time (Hobby), 60s (Pro)
1767
+ - Better for API routes than long-running processes
1768
+
1769
+ #### DigitalOcean Deployment
1770
+
1771
+ **Recommended: App Platform**
1772
+
1773
+ ```bash
1774
+ # Install doctl
1775
+ brew install doctl # or download from DigitalOcean
1776
+
1777
+ # Authenticate
1778
+ doctl auth init
1779
+
1780
+ # Create app from spec
1781
+ doctl apps create --spec .do/app.yaml
1782
+ ```
1783
+
1784
+ **Configuration** (`.do/app.yaml`):
1785
+ ```yaml
1786
+ name: llmguardian
1787
+ services:
1788
+ - name: api
1789
+ github:
1790
+ repo: dewitt4/LLMGuardian
1791
+ branch: main
1792
+ deploy_on_push: true
1793
+ dockerfile_path: docker/dockerfile
1794
+ http_port: 8000
1795
+ instance_count: 2
1796
+ instance_size_slug: professional-s
1797
+ routes:
1798
+ - path: /
1799
+ envs:
1800
+ - key: LOG_LEVEL
1801
+ value: INFO
1802
+ - key: ENVIRONMENT
1803
+ value: production
1804
+ health_check:
1805
+ http_path: /health
1806
+ ```
1807
+
1808
+ **Other DigitalOcean Options:**
1809
+ - **DOKS (DigitalOcean Kubernetes)**: Managed Kubernetes
1810
+ - **Droplets**: Traditional VMs with Docker
1811
+
1812
+ **Key Features:**
1813
+ - Simple pricing and scaling
1814
+ - Built-in monitoring
1815
+ - Automatic HTTPS
1816
+ - GitHub integration
1817
+
1818
+ ### Platform Comparison
1819
+
1820
+ | Feature | AWS | GCP | Azure | Vercel | DigitalOcean |
1821
+ |---------|-----|-----|-------|--------|--------------|
1822
+ | **Ease of Setup** | ⭐⭐⭐ | ⭐⭐⭐⭐⭐ | ⭐⭐⭐ | ⭐⭐⭐⭐⭐ | ⭐⭐⭐⭐ |
1823
+ | **Auto-Scaling** | Excellent | Excellent | Excellent | Automatic | Good |
1824
+ | **Cost (Monthly)** | $50-200 | $30-150 | $50-200 | $20-100 | $24-120 |
1825
+ | **Best For** | Enterprise | Startups | Enterprise | API/JAMstack | Simple Apps |
1826
+ | **Container Support** | ✅ ECS/EKS | ✅ Cloud Run/GKE | ✅ ACI/AKS | ❌ | ✅ App Platform |
1827
+ | **Serverless** | ✅ Lambda | ✅ Functions | ✅ Functions | ✅ Functions | Limited |
1828
+ | **Kubernetes** | ✅ EKS | ✅ GKE | ✅ AKS | ❌ | ✅ DOKS |
1829
+ | **Free Tier** | Yes | Yes | Yes | Yes | No |
1830
+
1831
+ ### Deployment Prerequisites
1832
+
1833
+ Before deploying to any cloud platform:
1834
+
1835
+ #### 1. Prepare Environment Configuration
1836
+
1837
+ ```bash
1838
+ # Copy and configure environment variables
1839
+ cp .env.example .env
1840
+
1841
+ # Edit with your settings
1842
+ nano .env
1843
+ ```
1844
+
1845
+ Key variables to set:
1846
+ - `SECURITY_RISK_THRESHOLD`
1847
+ - `API_SERVER_PORT`
1848
+ - `LOG_LEVEL`
1849
+ - `ENVIRONMENT` (production, staging, development)
1850
+ - API keys and secrets
1851
+
1852
+ #### 2. Build Docker Image
1853
+
1854
+ ```bash
1855
+ # Build from project root
1856
+ docker build -t llmguardian:latest -f docker/dockerfile .
1857
+
1858
+ # Test locally
1859
+ docker run -p 8000:8000 --env-file .env llmguardian:latest
1860
+ ```
1861
+
1862
+ #### 3. Set Up Cloud CLI Tools
1863
+
1864
+ **AWS:**
1865
+ ```bash
1866
+ # Install AWS CLI
1867
+ curl "https://awscli.amazonaws.com/awscli-exe-linux-x86_64.zip" -o "awscliv2.zip"
1868
+ unzip awscliv2.zip
1869
+ sudo ./aws/install
1870
+
1871
+ # Configure credentials
1872
+ aws configure
1873
+ ```
1874
+
1875
+ **GCP:**
1876
+ ```bash
1877
+ # Install gcloud SDK
1878
+ curl https://sdk.cloud.google.com | bash
1879
+ exec -l $SHELL
1880
+
1881
+ # Authenticate
1882
+ gcloud init
1883
+ gcloud auth login
1884
+ ```
1885
+
1886
+ **Azure:**
1887
+ ```bash
1888
+ # Install Azure CLI
1889
+ curl -sL https://aka.ms/InstallAzureCLIDeb | sudo bash
1890
+
1891
+ # Login
1892
+ az login
1893
+ ```
1894
+
1895
+ **Vercel:**
1896
+ ```bash
1897
+ # Install Vercel CLI
1898
+ npm i -g vercel
1899
+
1900
+ # Login
1901
+ vercel login
1902
+ ```
1903
+
1904
+ **DigitalOcean:**
1905
+ ```bash
1906
+ # Install doctl
1907
+ brew install doctl # macOS
1908
+ # or download from https://github.com/digitalocean/doctl
1909
+
1910
+ # Authenticate
1911
+ doctl auth init
1912
+ ```
1913
+
1914
+ #### 4. Configure Secrets Management
1915
+
1916
+ **AWS Secrets Manager:**
1917
+ ```bash
1918
+ aws secretsmanager create-secret \
1919
+ --name llmguardian-api-key \
1920
+ --secret-string "your-secret-key"
1921
+ ```
1922
+
1923
+ **GCP Secret Manager:**
1924
+ ```bash
1925
+ echo -n "your-secret-key" | gcloud secrets create llmguardian-api-key --data-file=-
1926
+ ```
1927
+
1928
+ **Azure Key Vault:**
1929
+ ```bash
1930
+ az keyvault create --name llmguardian-vault --resource-group llmguardian-rg
1931
+ az keyvault secret set --vault-name llmguardian-vault --name api-key --value "your-secret-key"
1932
+ ```
1933
+
1934
+ **Vercel:**
1935
+ ```bash
1936
+ vercel env add API_KEY
1937
+ # Enter secret when prompted
1938
+ ```
1939
+
1940
+ **DigitalOcean:**
1941
+ ```bash
1942
+ # Via App Platform dashboard or doctl
1943
+ doctl apps update YOUR_APP_ID --spec .do/app.yaml
1944
+ ```
1945
+
1946
+ ### Best Practices for Cloud Deployment
1947
+
1948
+ #### Security Hardening
1949
+
1950
+ 1. **Use Secret Managers**
1951
+ - Never hardcode secrets in code or environment files
1952
+ - Rotate secrets regularly
1953
+ - Use least-privilege IAM roles
1954
+
1955
+ 2. **Enable HTTPS/TLS**
1956
+ - Use cloud-provided certificates (free with most platforms)
1957
+ - Force HTTPS redirects
1958
+ - Configure SSL/TLS termination at load balancer
1959
+
1960
+ 3. **Implement WAF (Web Application Firewall)**
1961
+ - AWS: AWS WAF
1962
+ - Azure: Azure Application Gateway WAF
1963
+ - GCP: Cloud Armor
1964
+ - Vercel: Built-in DDoS protection
1965
+ - DigitalOcean: Cloud Firewalls
1966
+
1967
+ 4. **Network Security**
1968
+ - Configure VPCs/VNets for isolation
1969
+ - Use security groups/firewall rules
1970
+ - Implement least-privilege network policies
1971
+
1972
+ #### Monitoring & Logging
1973
+
1974
+ 1. **Enable Cloud-Native Monitoring**
1975
+ - AWS: CloudWatch
1976
+ - GCP: Cloud Monitoring & Logging
1977
+ - Azure: Application Insights
1978
+ - Vercel: Analytics
1979
+ - DigitalOcean: Built-in monitoring
1980
+
1981
+ 2. **Configure Alerts**
1982
+ ```bash
1983
+ # Example: AWS CloudWatch alarm
1984
+ aws cloudwatch put-metric-alarm \
1985
+ --alarm-name llmguardian-high-cpu \
1986
+ --alarm-description "Alert when CPU exceeds 80%" \
1987
+ --metric-name CPUUtilization \
1988
+ --threshold 80
1989
+ ```
1990
+
1991
+ 3. **Set Up Log Aggregation**
1992
+ - Centralize logs for analysis
1993
+ - Implement log retention policies
1994
+ - Enable audit logging
1995
+
1996
+ #### Performance Optimization
1997
+
1998
+ 1. **Auto-Scaling Configuration**
1999
+ - Set appropriate min/max instances
2000
+ - Configure based on CPU/memory metrics
2001
+ - Implement graceful shutdown
2002
+
2003
+ 2. **Caching**
2004
+ - Use Redis/Memcached for response caching
2005
+ - Implement CDN for static content
2006
+ - Cache embeddings and common queries
2007
+
2008
+ 3. **Database Optimization**
2009
+ - Use managed database services
2010
+ - Implement connection pooling
2011
+ - Regular performance monitoring
2012
+
2013
+ #### Cost Optimization
2014
+
2015
+ 1. **Right-Sizing**
2016
+ - Start small and scale based on metrics
2017
+ - Use spot/preemptible instances for non-critical workloads
2018
+ - Monitor and optimize resource usage
2019
+
2020
+ 2. **Reserved Instances**
2021
+ - Purchase reserved capacity for predictable workloads
2022
+ - 1-year or 3-year commitments for savings
2023
+
2024
+ 3. **Cost Alerts**
2025
+ ```bash
2026
+ # AWS Budget alert
2027
+ aws budgets create-budget \
2028
+ --account-id YOUR_ACCOUNT_ID \
2029
+ --budget file://budget.json
2030
+ ```
2031
+
2032
+ ### CI/CD Integration
2033
+
2034
+ **GitHub Actions Example** (`.github/workflows/deploy-cloud.yml`):
2035
+
2036
+ ```yaml
2037
+ name: Deploy to Cloud
2038
+
2039
+ on:
2040
+ push:
2041
+ branches: [main]
2042
+ workflow_dispatch:
2043
+
2044
+ jobs:
2045
+ deploy-aws:
2046
+ runs-on: ubuntu-latest
2047
+ steps:
2048
+ - uses: actions/checkout@v4
2049
+
2050
+ - name: Configure AWS credentials
2051
+ uses: aws-actions/configure-aws-credentials@v1
2052
+ with:
2053
+ aws-access-key-id: ${{ secrets.AWS_ACCESS_KEY_ID }}
2054
+ aws-secret-access-key: ${{ secrets.AWS_SECRET_ACCESS_KEY }}
2055
+ aws-region: us-east-1
2056
+
2057
+ - name: Login to Amazon ECR
2058
+ run: |
2059
+ aws ecr get-login-password --region us-east-1 | docker login --username AWS --password-stdin ${{ secrets.AWS_ACCOUNT_ID }}.dkr.ecr.us-east-1.amazonaws.com
2060
+
2061
+ - name: Build and push
2062
+ run: |
2063
+ docker build -t llmguardian:latest -f docker/dockerfile .
2064
+ docker tag llmguardian:latest ${{ secrets.AWS_ACCOUNT_ID }}.dkr.ecr.us-east-1.amazonaws.com/llmguardian:latest
2065
+ docker push ${{ secrets.AWS_ACCOUNT_ID }}.dkr.ecr.us-east-1.amazonaws.com/llmguardian:latest
2066
+
2067
+ - name: Deploy to ECS
2068
+ run: |
2069
+ aws ecs update-service --cluster llmguardian-cluster --service llmguardian-service --force-new-deployment
2070
+
2071
+ deploy-gcp:
2072
+ runs-on: ubuntu-latest
2073
+ steps:
2074
+ - uses: actions/checkout@v4
2075
+
2076
+ - uses: google-github-actions/setup-gcloud@v0
2077
+ with:
2078
+ service_account_key: ${{ secrets.GCP_SA_KEY }}
2079
+ project_id: ${{ secrets.GCP_PROJECT_ID }}
2080
+
2081
+ - name: Deploy to Cloud Run
2082
+ run: |
2083
+ gcloud auth configure-docker
2084
+ docker build -t llmguardian:latest -f docker/dockerfile .
2085
+ docker tag llmguardian:latest gcr.io/${{ secrets.GCP_PROJECT_ID }}/llmguardian:latest
2086
+ docker push gcr.io/${{ secrets.GCP_PROJECT_ID }}/llmguardian:latest
2087
+ gcloud run deploy llmguardian --image gcr.io/${{ secrets.GCP_PROJECT_ID }}/llmguardian:latest --region us-central1
2088
+
2089
+ deploy-azure:
2090
+ runs-on: ubuntu-latest
2091
+ steps:
2092
+ - uses: actions/checkout@v4
2093
+
2094
+ - uses: azure/login@v1
2095
+ with:
2096
+ creds: ${{ secrets.AZURE_CREDENTIALS }}
2097
+
2098
+ - name: Deploy to Azure
2099
+ run: |
2100
+ az acr login --name llmguardianacr
2101
+ docker build -t llmguardian:latest -f docker/dockerfile .
2102
+ docker tag llmguardian:latest llmguardianacr.azurecr.io/llmguardian:latest
2103
+ docker push llmguardianacr.azurecr.io/llmguardian:latest
2104
+ az container restart --resource-group llmguardian-rg --name llmguardian-container
2105
+ ```
2106
+
2107
+ ### Troubleshooting Common Issues
2108
+
2109
+ #### Port Binding Issues
2110
+ ```bash
2111
+ # Ensure correct port exposure
2112
+ docker run -p 8000:8000 llmguardian:latest
2113
+
2114
+ # Check health endpoint
2115
+ curl http://localhost:8000/health
2116
+ ```
2117
+
2118
+ #### Memory/CPU Limits
2119
+ ```bash
2120
+ # Increase container resources
2121
+ # AWS ECS: Update task definition
2122
+ # GCP Cloud Run: Use --memory and --cpu flags
2123
+ # Azure: Update container instance specs
2124
+ ```
2125
+
2126
+ #### Environment Variables Not Loading
2127
+ ```bash
2128
+ # Verify environment variables
2129
+ docker run llmguardian:latest env | grep LOG_LEVEL
2130
+
2131
+ # Check cloud secret access
2132
+ # AWS: Verify IAM role permissions
2133
+ # GCP: Check service account permissions
2134
+ # Azure: Verify Key Vault access policies
2135
+ ```
2136
+
2137
+ #### Image Pull Failures
2138
+ ```bash
2139
+ # Authenticate with registry
2140
+ aws ecr get-login-password | docker login --username AWS --password-stdin YOUR_REGISTRY
2141
+ gcloud auth configure-docker
2142
+ az acr login --name YOUR_REGISTRY
2143
+ ```
2144
+
2145
+ ### Additional Resources
2146
+
2147
+ - **[PROJECT.md - Complete Cloud Deployment Guides](../PROJECT.md#cloud-deployment-guides)**: Full step-by-step instructions with all configuration details
2148
+ - **[Docker README](../docker/README.md)**: Docker-specific documentation
2149
+ - **[Environment Variables](.env.example)**: All configuration options
2150
+ - **[GitHub Actions Workflows](../.github/workflows/README.md)**: CI/CD automation
2151
+
2152
+ ### Support
2153
+
2154
+ For deployment issues:
2155
+ 1. Check the [GitHub Issues](https://github.com/dewitt4/LLMGuardian/issues)
2156
+ 2. Review cloud provider documentation
2157
+ 3. Enable debug logging: `LOG_LEVEL=DEBUG`
2158
+ 4. Check health endpoint: `curl http://your-deployment/health`
2159
+
2160
+ ---
2161
+
2162
+ **Ready to deploy? Choose your platform above and follow the deployment guide!** 🚀