Skip to main content

​
Overview

By default, workloads in DCDeploy use ephemeral storage — data is lost when the container restarts, redeploys, or scales.
To store data reliably across restarts and deployments, DCDeploy provides Persistent Volumes. Persistent volumes are ideal for databases, file storage, or any workload that needs durable storage.

​
Use Cases

  • Running databases like Postgres, MySQL, MongoDB.
  • Storing uploaded files (images, PDFs, media).
  • Maintaining caches, logs, or stateful services that must survive restarts.

​
Prerequisites

  • An active environment in DCDeploy.
  • A workload with volume support enabled.
  • Sufficient quota for storage (allocated at org level).

​
Step-by-Step Guide

​
1. Create a Service with Volume

In the DCDeploy dashboard:
  1. Go to Deploy → Add Service.
  2. Under Workload Settings, choose Persistent Volume.
  3. Set the volume size (e.g., 10GB).
  4. Mount it at a path inside the container (e.g., /data).
Example YAML (DCDeploy.yml): 
services:
  my-db:
    image: postgres:15
    volumes:
      - mountPath: /var/lib/postgresql/data
        size: 20GB

​
2. Access the Volume

Inside your container, the mounted path behaves like a local folder. Example (Postgres): 
psql -h localhost -U user -d mydb
# Data will persist in /var/lib/postgresql/data across restarts

​
3. Scaling Behavior

  • Vertical scaling (CPU/RAM changes): volume is preserved.
  • Horizontal scaling (multiple instances): each instance gets its own volume (data is not automatically shared).
  • For shared storage, use an external database or object storage.

​
Best Practices

  • Always mount databases to a persistent volume.
  • Avoid storing large assets directly in volumes; use object storage (S3, GCS) instead.
  • Use backup tools to periodically snapshot data.
  • Keep volumes lightweight for faster redeploys.

​
Troubleshooting

  • Data loss after redeploy: Ensure you’re using a persistent volume, not ephemeral storage.
  • Multiple replicas with inconsistent data: Volumes are not shared — use an external DB if consistency is required.
  • Storage full errors: Increase the volume size via dashboard or update your DCDeploy.yml.