Upgrade Guide

How to upgrade Sligo Enterprise to a new version.

Before Upgrading

1. Check Release Notes

Review release notes for:

  • Breaking changes
  • New features
  • Migration steps required
  • Deprecated features

2. Backup

For in-cluster database:

# Backup PostgreSQL
kubectl exec postgres-0 -n sligo -- pg_dump -U sligo sligo > backup-$(date +%Y%m%d).sql

For external database (RDS):

  • Take RDS snapshot via AWS console

3. Review Current Configuration

# Get current values
helm get values sligo-app -n sligo > current-values.yaml

# Check current version
helm list -n sligo

Standard Upgrade Process

1. Update Helm Repository

helm repo update sligo

2. Check Available Versions

helm search repo sligo/sligo-cloud --versions

3. Review Changes

# See what will change
helm diff upgrade sligo-app sligo/sligo-cloud \
  --version NEW_VERSION \
  -f values-production.yaml \
  -n sligo

4. Perform Upgrade

helm upgrade sligo-app sligo/sligo-cloud \
  --version NEW_VERSION \
  -f values-production.yaml \
  -n sligo \
  --timeout 10m

5. Verify Upgrade

# Check rollout status
kubectl rollout status deployment/sligo-app -n sligo
kubectl rollout status deployment/sligo-backend -n sligo
kubectl rollout status deployment/mcp-gateway -n sligo

# Check pods
kubectl get pods -n sligo

# Check application health
curl https://app.your-domain.com/api/health

Rollback

If upgrade fails or causes issues:

# Rollback to previous version
helm rollback sligo-app -n sligo

# Or rollback to specific revision
helm rollback sligo-app 1 -n sligo

# Check history
helm history sligo-app -n sligo

Version-Specific Upgrade Notes

Upgrading to 1.1.0 from 1.0.x

Changes:

  • Added autoscaling support
  • New ingress annotations for AWS ALB

Migration steps:

  • Review new autoscaling values
  • Update ALB annotations if using AWS
  • No breaking changes

Upgrading to 2.0.0 from 1.x.x

Breaking changes:

  • Secrets structure changed
  • Database configuration restructured

Migration steps:

  1. Update secrets: ```bash

    Old format (deprecated)

    kubectl delete secret app-secrets -n sligo

New format

kubectl create secret generic nextjs-secrets
–from-literal=DATABASE_URL=”…”
…


2. **Update values.yaml:**
```yaml
# Old format
database:
  host: postgres

# New format
database:
  type: internal
  internal:
    name: postgres
  1. Perform upgrade

Zero-Downtime Upgrades

For production systems requiring zero downtime:

1. Use Rolling Updates (Default)

Helm automatically performs rolling updates:

# In values.yaml
app:
  replicaCount: 3  # Must be > 1 for zero downtime

2. Blue-Green Deployment (Advanced)

# Install new version with different release name
helm install sligo-app-v2 sligo/sligo-cloud \
  --version NEW_VERSION \
  -f values-production.yaml \
  -n sligo

# Switch traffic via ingress update
# Then remove old version:
helm uninstall sligo-app -n sligo

3. Canary Deployment (Advanced)

Use tools like Flagger or Argo Rollouts for gradual traffic shifting.

Database Migrations

If upgrade includes database schema changes:

1. Pre-Upgrade Migration

# Run migration job before upgrade
kubectl apply -f migration-job.yaml -n sligo

# Wait for completion
kubectl wait --for=condition=complete job/db-migration -n sligo --timeout=300s

2. Post-Upgrade Verification

# Verify database schema
kubectl exec -it postgres-0 -n sligo -- psql -U sligo -d sligo -c "\dt"

Configuration Updates

Adding New Features

# In values-production.yaml

# Enable autoscaling
app:
  autoscaling:
    enabled: true
    minReplicas: 2
    maxReplicas: 10

Apply changes:

helm upgrade sligo-app sligo/sligo-cloud \
  -f values-production.yaml \
  -n sligo

Updating Resource Limits

app:
  resources:
    limits:
      cpu: 1000m  # Updated from 500m
      memory: 1Gi  # Updated from 512Mi

Updating Image Tags

To upgrade to a new application version:

  1. Check available versions in Google Artifact Registry or contact Sligo support
  2. Update image tags in your values.yaml: 
    app:
  image:
    tag: "v1.2.0"  # Updated from v1.1.0
   
backend:
  image:
    tag: "v1.2.0"  # Updated from v1.1.0
   
mcpGateway:
  image:
    tag: "v1.2.0"  # Updated from v1.1.0
  3. Apply the upgrade: 
    helm upgrade sligo-app sligo/sligo-cloud \
  -f values-production.yaml \
  -n sligo

Important: Always use version tags (e.g., v1.0.0, v1.2.3) in production. Avoid using latest as it can cause unexpected updates.

Troubleshooting Upgrades

Upgrade Stuck

# Check status
helm status sligo-app -n sligo

# Check pods
kubectl get pods -n sligo -w

# Check events
kubectl get events -n sligo --sort-by='.lastTimestamp'

Rollback Failed

If automatic rollback fails:

# Force delete stuck resources
kubectl delete pod <stuck-pod> -n sligo --grace-period=0 --force

# Manual rollback
helm rollback sligo-app -n sligo --force

Configuration Errors

# Validate new values
helm template test sligo/sligo-cloud \
  -f values-production.yaml \
  --debug

# Check differences
helm diff upgrade sligo-app sligo/sligo-cloud \
  -f values-production.yaml \
  -n sligo

Best Practices

  • Always test upgrades in staging first
  • Backup before upgrading
  • Review release notes for breaking changes
  • Monitor application after upgrade
  • Keep old backups for at least 30 days
  • Document any customizations made
  • Use version pinning in production
# In values-production.yaml
app:
  image:
    tag: "v1.2.0"  # Pin specific version, not "latest"

Scheduled Maintenance

Plan upgrades during low-traffic periods:

  • Notify users of maintenance window
  • Prepare rollback plan
  • Have support team on standby
  • Monitor metrics during and after upgrade
  • Keep communication channels open

Support

For upgrade assistance:

  • Email: support@sligo.ai
  • Include: Current version, target version, error messages
Edit this page on GitHub