Advanced Workflows

Complex multi-step automation scenarios with Qovery AI Copilot

Multi-Step Deployments

Progressive Deployment Pipeline

You: "Deploy the new version to dev, then staging, then production - but only if tests pass at each stage"

Copilot:
1. Deploys to dev environment
2. Runs automated tests
3. If tests pass, deploys to staging
4. Runs staging tests
5. Waits for approval before production
6. Deploys to production
7. Monitors production health

Blue-Green Deployment

You: "Set up a blue-green deployment for my-api in production"

Copilot:
1. Creates green environment (copy of blue)
2. Deploys new version to green
3. Runs health checks on green
4. Switches traffic to green
5. Monitors for issues
6. Keeps blue as backup

Canary Deployment

You: "Deploy version 2.0 as a canary to 10% of production traffic"

Copilot:
1. Deploys new version alongside current
2. Configures traffic split (90% old, 10% new)
3. Monitors error rates
4. Gradually increases traffic if healthy
5. Rolls back if errors spike

Complex Infrastructure Changes

Multi-Environment Setup

You: "Create a complete staging infrastructure that mirrors production but with half the resources"

Copilot:
1. Analyzes production infrastructure
2. Creates staging environment
3. Clones all services
4. Reduces resource allocations by 50%
5. Adjusts database sizes
6. Configures networking
7. Sets up monitoring
8. Validates complete setup

Infrastructure Migration

You: "Migrate all staging environments from cluster-east to cluster-west"

Copilot:
1. Lists all staging environments
2. For each environment:
   - Creates new environment in cluster-west
   - Deploys all services
   - Migrates data
   - Validates functionality
   - Switches DNS/routing
   - Deletes old environment
3. Reports migration summary

Time-Based Automation

Scheduled One-Time Tasks

You: "Stop all development environments tonight at 8 PM"

Copilot:
1. Identifies all development environments
2. Schedules shutdown task for 8 PM
3. Confirms scheduled action
4. Executes at specified time
5. Sends notification when complete

You: "Deploy the hotfix to production tomorrow at 2 AM"

Copilot:
1. Validates hotfix is ready
2. Schedules deployment for 2 AM
3. Sets up pre-deployment checks
4. Executes deployment at scheduled time
5. Monitors post-deployment health
6. Alerts team of completion

Recurring Scheduled Tasks

You: "Every weekday at 6 PM, stop all dev and staging environments"

Copilot:
1. Creates recurring schedule (Mon-Fri at 6 PM)
2. Identifies target environments
3. Configures automatic shutdown
4. Sends daily summary reports
5. Maintains schedule until cancelled

You: "Start all development environments every morning at 8 AM on weekdays"

Copilot:
1. Sets up recurring schedule (Mon-Fri at 8 AM)
2. Lists all dev environments
3. Configures automatic startup
4. Ensures environments are ready by 8:30 AM
5. Notifies team when environments are available

You: "Run database backups every night at 2 AM and keep for 7 days"

Copilot:
1. Creates nightly backup schedule (2 AM daily)
2. Configures backup retention (7 days)
3. Sets up backup verification
4. Monitors backup success/failure
5. Alerts on backup issues
6. Manages automatic cleanup of old backups

Time-Based Cost Optimization

You: "Automatically scale down all non-production services to minimum replicas on weekends"

Copilot:
1. Identifies non-production environments
2. Creates weekend schedule (Fri 6 PM to Mon 8 AM)
3. Records current replica counts
4. Scales to minimum on Friday evening
5. Restores original counts Monday morning
6. Reports weekly cost savings

You: "During business hours (9 AM - 6 PM), ensure production has 5 replicas. Outside hours, scale to 2 replicas."

Copilot:
1. Sets up two recurring schedules
2. Scale up schedule: Weekdays 9 AM (to 5 replicas)
3. Scale down schedule: Weekdays 6 PM (to 2 replicas)
4. Monitors application performance during transitions
5. Tracks cost savings from reduced off-hours capacity

Maintenance Windows

You: "Every Sunday from 2 AM to 4 AM, perform maintenance on staging: update all services, run database optimization, clear caches"

Copilot:
1. Creates weekly maintenance window (Sun 2-4 AM)
2. Schedules sequential tasks:
   - 2:00 AM: Deploy latest versions to all services
   - 2:30 AM: Run database optimization scripts
   - 3:00 AM: Clear application caches
   - 3:30 AM: Run health checks
3. Monitors each task completion
4. Sends maintenance report at 4 AM
5. Alerts if any task fails

Conditional Time-Based Actions

You: "If CPU usage stays above 80% for more than 30 minutes during business hours, scale up. After 6 PM, just alert me instead."

Copilot:
1. Sets up time-based conditional rules:
   - Business hours (9 AM - 6 PM): Auto-scale on high CPU
   - After hours (6 PM - 9 AM): Alert only
2. Monitors CPU metrics continuously
3. During business hours:
   - Detects sustained high CPU
   - Automatically scales up
   - Notifies of action taken
4. After hours:
   - Detects sustained high CPU
   - Sends alert without scaling
   - Waits for manual intervention

Ephemeral Environment Lifecycle

You: "Create a preview environment for PR #123 and automatically delete it in 7 days or when the PR is merged"

Copilot:
1. Creates preview environment for PR #123
2. Deploys PR branch
3. Sets up two deletion triggers:
   - Time-based: Auto-delete after 7 days
   - Event-based: Auto-delete on PR merge
4. Monitors PR status
5. Deletes environment when either condition is met
6. Notifies team of deletion

You: "Every Monday, delete all feature environments older than 2 weeks"

Copilot:
1. Creates weekly cleanup schedule (Mondays)
2. Scans for feature environments
3. Identifies environments older than 2 weeks
4. Checks for recent activity
5. Deletes inactive old environments
6. Sends weekly cleanup report
7. Reports cost savings from cleanup

Automated Optimization

Cost Optimization Workflow

You: "Analyze my infrastructure and implement cost savings where possible"

Copilot:
1. Analyzes all environments
2. Identifies underutilized resources
3. Finds idle services
4. Suggests optimizations:
   - Right-size over-provisioned services
   - Stop inactive environments
   - Use spot instances where appropriate
   - Consolidate low-traffic services
5. Implements approved optimizations
6. Reports projected monthly savings

Performance Optimization

You: "Optimize the performance of my production stack"

Copilot:
1. Analyzes performance metrics
2. Identifies bottlenecks
3. Reviews configurations
4. Suggests improvements:
   - Increase resources for constrained services
   - Optimize database queries
   - Add caching layers
   - Adjust auto-scaling rules
5. Implements approved changes
6. Monitors impact

Disaster Recovery

Automated Backup and Restore

You: "Backup production, make breaking changes to staging, then restore if needed"

Copilot:
1. Takes snapshot of production state
2. Applies changes to staging
3. Monitors for issues
4. If problems detected:
   - Automatically restores from snapshot
   - Documents what went wrong
   - Suggests fixes

Multi-Region Failover

You: "If production goes down, automatically failover to DR region"

Copilot:
1. Monitors production health
2. If failure detected:
   - Validates DR environment is healthy
   - Switches DNS to DR region
   - Notifies team
   - Provides status dashboard
3. When production recovers:
   - Syncs data
   - Fails back to primary

Batch Operations

Bulk Environment Management

You: "For all feature branches older than 2 weeks: stop them if inactive, delete them if older than 1 month"

Copilot:
1. Lists all feature branch environments
2. Checks last activity for each
3. Categorizes by age and activity
4. Stops inactive 2-week-old environments
5. Deletes 1-month-old environments
6. Generates cleanup report
7. Calculates cost savings

Mass Configuration Updates

You: "Update the API_VERSION environment variable to 2.0 across all environments"

Copilot:
1. Lists all environments
2. For each environment:
   - Finds services with API_VERSION
   - Updates value to 2.0
   - Triggers redeployment
   - Waits for healthy status
3. Reports success/failures

CI/CD Integration

Automated Release Pipeline

You: "When a new tag is pushed, automatically deploy to staging, run tests, and notify Slack"

Copilot:
1. Sets up webhook for new tags
2. On tag event:
   - Deploys to staging
   - Waits for deployment
   - Runs integration tests
   - If tests pass:
     - Notifies Slack with success
     - Creates production deployment plan
   - If tests fail:
     - Rolls back staging
     - Notifies Slack with failure details
     - Logs for review

Hotfix Workflow

You: "I need to deploy a hotfix immediately - bypass normal pipeline"

Copilot:
1. Identifies hotfix branch
2. Creates emergency deployment plan
3. Deploys directly to production
4. Monitors closely for issues
5. Documents hotfix deployment
6. Reminds to backport to main

Data Migration

Zero-Downtime Migration

You: "Migrate the database from PostgreSQL 13 to 15 with zero downtime"

Copilot:
1. Creates new PostgreSQL 15 database
2. Sets up replication from old to new
3. Monitors replication lag
4. When in sync:
   - Switches application to new database
   - Monitors for errors
   - Keeps old database as backup
5. After validation period:
   - Removes old database
   - Updates documentation

Data Archival

You: "Archive all data from environments deleted more than 3 months ago"

Copilot:
1. Identifies deleted environments
2. Locates associated data
3. Creates archive storage
4. Exports data to archive
5. Verifies archive integrity
6. Removes from active storage
7. Updates inventory

Compliance & Auditing

Compliance Scan

You: "Audit all production services for security compliance"

Copilot:
1. Scans all production services
2. Checks:
   - Environment variables not hardcoded
   - No public database access
   - HTTPS enforced
   - Secrets properly managed
   - Latest security patches applied
3. Generates compliance report
4. Flags violations
5. Suggests remediations

Access Review

You: "Review who has access to production and ensure least privilege"

Copilot:
1. Lists all users with production access
2. Reviews access levels
3. Identifies overly permissive access
4. Suggests access reductions
5. Generates access audit report

Monitoring & Alerting

Proactive Issue Detection

You: "Monitor production and automatically fix common issues"

Copilot:
1. Monitors production continuously
2. Detects patterns like:
   - Memory leaks (restart affected services)
   - Connection pool exhaustion (scale up)
   - Failed health checks (redeploy)
3. Automatically remediates known issues
4. Notifies of actions taken
5. Escalates unknown issues

Intelligent Alerting

You: "Alert me only for critical issues, auto-handle everything else"

Copilot:
1. Monitors all infrastructure
2. Categorizes issues by severity
3. Auto-remediates:
   - Low: Service restarts, scaling adjustments
   - Medium: Failovers, traffic rerouting
4. Alerts for:
   - High: Data corruption, security issues
   - Critical: Complete outages, data breaches
5. Provides context and suggested actions

Best Practices for Complex Workflows

Break Down Complex Requests

For very complex workflows, break them into phases:

1. “First, set up the infrastructure…”
2. “Next, configure the services…”
3. “Finally, validate everything…”

Use Checkpoints

Request confirmation at critical points:

“Deploy to staging, wait for my approval, then continue to production”

Define Success Criteria

Be explicit about what success looks like:

“Deploy and consider it successful only if error rate stays below 0.1%”

Plan for Failures

Include failure handling in your requests:

“Try to deploy, but if it fails, automatically rollback and notify me”

Next Steps

Common Tasks

Start with simpler everyday operations

Infrastructure Management

Learn infrastructure operations

Deployment Automation

Master deployment capabilities

Optimization

Optimize your workflows