Disaster Recovery Manual

GCP Cold DR Runbook with XReplicator

Full-width runbook for recovering Compute Engine workloads from XReplicator snapshots using boot-disk snapshot to image conversion, DR instance creation, and restored data-disk attachment.

Workflow

  1. Restore boot and data disks to a staging host with XReplicator disk restore.
  2. Create persistent disk snapshot from restored boot disk.
  3. Create image from that snapshot.
  4. Create DR VM from image.
  5. Attach restored data disks and validate application readiness.

Prerequisites

  • Backup server reachable from DR VPC/project.
  • Staging Compute Engine VM with XReplicator agent connected.
  • Target Persistent Disks attached with same sizes as source disks.
  • No pre-formatting required on restored target disks.

CLI Commands

Run after XReplicator has restored source disk content to staging-attached disks.

Set variables

# Required variables
export PROJECT_ID=my-dr-project
export ZONE=us-central1-a
export MACHINE_TYPE=n2-standard-4
export NETWORK=default
export SUBNET=default
export SERVICE_ACCOUNT=dr-vm@my-dr-project.iam.gserviceaccount.com
export DR_VM=dr-recovered-vm

# Restored disks from XReplicator on staging VM
export RESTORED_BOOT_DISK=restored-boot-disk
export RESTORED_DATA_DISK_1=restored-data-disk-1
export RESTORED_DATA_DISK_2=restored-data-disk-2

gcloud config set project "$PROJECT_ID"

Create boot snapshot

# 1) Create snapshot from restored boot disk
BOOT_SNAP=dr-boot-snap-$(date +%Y%m%d%H%M%S)
gcloud compute disks snapshot "$RESTORED_BOOT_DISK" \
  --zone "$ZONE" \
  --snapshot-names "$BOOT_SNAP"

echo "Boot snapshot: $BOOT_SNAP"

Snapshot to image conversion

# 2) Create image from snapshot
DR_IMAGE=dr-boot-image-$(date +%Y%m%d%H%M%S)
gcloud compute images create "$DR_IMAGE" \
  --source-snapshot "$BOOT_SNAP" \
  --family dr-recovery-family

echo "DR image: $DR_IMAGE"

Bring up DR VM from image

# 3) Bring up DR VM from recovered image
gcloud compute instances create "$DR_VM" \
  --zone "$ZONE" \
  --machine-type "$MACHINE_TYPE" \
  --subnet "$SUBNET" \
  --image "$DR_IMAGE" \
  --service-account "$SERVICE_ACCOUNT" \
  --scopes https://www.googleapis.com/auth/cloud-platform

Attach restored data disks

# 4) Detach restored data disks from staging VM (if still attached there)
# Replace STAGING_VM with your staging host name
gcloud compute instances detach-disk STAGING_VM \
  --disk "$RESTORED_DATA_DISK_1" \
  --zone "$ZONE"

gcloud compute instances detach-disk STAGING_VM \
  --disk "$RESTORED_DATA_DISK_2" \
  --zone "$ZONE"

# 5) Attach restored data disks to DR VM
gcloud compute instances attach-disk "$DR_VM" \
  --disk "$RESTORED_DATA_DISK_1" \
  --zone "$ZONE"

gcloud compute instances attach-disk "$DR_VM" \
  --disk "$RESTORED_DATA_DISK_2" \
  --zone "$ZONE"

Validation Checklist

  • Recovered VM boots from new image.
  • All restored data disks are attached and mounted.
  • Application and service checks pass.
  • Traffic cutover and DNS updates are completed.
Back to Disaster RecoveryBack to Solutions