Dgraph Backup and Restore

This article provides a detailed walkthrough for managing Dgraph backups and restoration within your on-premises environment. Proper configuration of the stitchit service ensures that graph data is preserved and recoverable in the event of a system failure or data loss scenario.

This guide provides detailed procedures for backup and restore of Dgraph in on-premises Kubernetes Provider Connectivity Assurance deployments.

Prerequisites

kubectl access to the Provider Connectivity Assurance cluster
SSH access to cluster nodes (for restoration)
Familiarity with OpenEBS local storage paths

Backup Overview

Dgraph backups are managed by the stitchit service using configuration from the stitchit-env-config ConfigMap. Unlike other components that use CronJobs, Dgraph backup scheduling is controlled through this ConfigMap.

View the backup configuration:

kubectl describe configmap -n pca stitchit-env-config

Key configuration values:

Setting	Description
`S3_BUCKET`	MinIO bucket path for backups
`S3_SERVER`	MinIO server endpoint
`BACKUP_SCHEDULE`	Cron expression for backup timing
`BACKUP_RETENTION`	Retention period (e.g., `P15D` = 15 days)

Default behavior:

Runs daily at 04:00 UTC
Writes backups to the local MinIO storage backend
Retains backups for 15 days

Manual Backup Trigger

You can trigger a Dgraph backup using the GraphQL admin API from any pod with curl:

kubectl exec -it -n pca airflow-0 -- sh

Retrieve the MinIO credentials:

kubectl exec -it -n pca pca-minio-pool-0-0 -- cat /tmp/minio/config.env | grep ROOT

Execute the backup mutation:

curl -H "Content-Type: application/graphql" -X POST dgraphalpha1-0:8080/admin -d '
mutation {
  export(input: {
    destination: "minio://pca-minio-hl.pca.svc.cluster.local:9000/<bucket-name>?secure=false"
    accessKey: "<MINIO_ROOT_USER>"
    secretKey: "<MINIO_ROOT_PASSWORD>"
  }) {
    response {
      message
      code
    }
  }
}
'

Replace <bucket-name>, <MINIO_ROOT_USER>, and <MINIO_ROOT_PASSWORD> with your actual values.

Alternative: Modify Backup Schedule

You can trigger an immediate backup by temporarily changing the schedule:

Edit the ConfigMap:

kubectl edit configmap -n pca stitchit-env-config

Change BACKUP_SCHEDULE to a time a few minutes in the future.

Restart the stitchit deployment:

kubectl rollout restart deploy -n pca stitchit

After the backup completes, restore the original schedule.

Access Backups in MinIO

Connect to the MinIO pod:

kubectl exec -it -n pca pca-minio-pool-0-0 -- sh
. /tmp/minio/config.env

Configure the MinIO client:

mc alias set --insecure pca https://localhost:9000 "$MINIO_ROOT_USER" "$MINIO_ROOT_PASSWORD"

List available backups:

mc --insecure ls pca/<bucket-name>/dgraphBackup/v2/

Copy a backup to the pod filesystem:

mc --insecure cp -r \
  pca/<bucket-name>/dgraphBackup/v2/<backup-folder>/ \
  /tmp/

Restore Procedure

Important: Dgraph restoration procedures are complex and depend on your specific deployment configuration. Contact Cisco TAC for guidance on restoring Dgraph data in your environment.

The general restoration approach involves:

Export the backup from MinIO to the admin VM
Stop Dgraph services
Clear existing Dgraph data
Import the backup using Dgraph's bulk loader or live loader
Restart services and validate

Post-Restore Validation

Verify Dgraph cluster health
Confirm data queries return expected results
Check dependent services are functioning correctly