Virtuoso SaaS Managed Service Buyer Backup and Restore Guide

Virtuoso Universal Server
, ,
1

Virtuoso SaaS Managed Service Buyer Backup and Restore Guide

This guide explains the backup operations available to Virtuoso SaaS buyers and
how to request a controlled restore from OpenLink.

Scope

Buyers can:

  • Run an online Virtuoso backup from the buyer portal.
  • Confirm that the backup-sync operation completed successfully.
  • Retain the reported backup date, time, and durable S3 prefix for a future
    support request.
  • Request a seller-admin controlled restore of a selected backup set.

Buyers cannot run a restore directly. A restore replaces the active database
state, requires a service outage, and is performed only by OpenLink seller
administrators after explicit buyer authorization.

Before You Start

You need:

The portal provides a tenant-specific script. Do not reuse a script from
another buyer, instance, or environment.

Run an Online Backup

  1. Sign in to the Virtuoso SaaS Buyer Portal.
  2. Open Instances and select the running instance.
  3. Open Backup & Recovery.
  4. Select Copy Script or Download Script for Run online backup.
  5. Run the generated script in a terminal configured with the buyer IAM user
    or role.
  6. Wait for the script to complete. Do not close the terminal while the
    backup-sync task is running.

The generated script:

  1. Assumes the buyer’s scoped ECS Exec role for that instance.
  2. Configures the Virtuoso backup directory.
  3. Runs backup_online('virtuoso_#', ...) while Virtuoso remains online.
  4. Starts a scoped Fargate sync task to copy the backup set to durable S3.
  5. Waits for the sync task and reports success only when its exit code is 0.

The first successful run establishes a full backup chain. The service then runs
its per-instance daily backup workflow, which creates incremental backups and
copies them to the same durable S3 location.

Verify the Backup

A successful script finishes without an error and reports the durable S3 backup
prefix for the instance. Record:

  • Instance ID and project name.
  • Completion date and UTC time.
  • Backup prefix or backup-set timestamp.
  • Script output confirming the S3 sync completed successfully.

The S3 bucket is managed by OpenLink. Buyers should not alter, delete, or
replace objects in the managed backup prefix. If independent S3 object
verification is required, request it from OpenLink support and include the
instance ID and reported backup prefix.

Request a Restore

A restore is not an online operation. It replaces the active Virtuoso database
with a selected backup set and can discard changes made after that backup.

Submit a support request that includes:

  • Buyer AWS account ID and buyer portal email address.
  • Instance ID and project name.
  • The backup date/time or backup-set prefix to restore.
  • Explicit authorization to replace the current database state.
  • Confirmation that the expected service outage is acceptable.
  • A technical contact to confirm post-restore validation.

OpenLink will:

  1. Verify the buyer, instance ownership, and requested backup set.
  2. Create a rollback copy of the current database files.
  3. Stop the tenant service and restore the selected backup set.
  4. Restart Virtuoso and verify that the service becomes healthy.
  5. Confirm the result with the buyer and retain the most recent rollback copy
    for the configured retention period.

Do not delete the instance to request a restore. Instance deletion is a
termination action and is separate from backup recovery.

Validate After a Restore

After OpenLink confirms completion:

  1. Open the instance in the Virtuoso SaaS Buyer Portal.
  2. Confirm status is running and health is healthy.
  3. Open HTTPS Conductor and verify the expected database content.
  4. Run an application-level or SQL validation appropriate for the restored
    workload.
  5. Run a new online backup after validation if the restored state must be
    protected immediately.

Troubleshooting

The Backup & Recovery section is missing

The instance may still be provisioning, stopped, or missing required backup
metadata. Confirm that it is running and healthy; if the section does not
appear, contact OpenLink support with the instance ID.

The backup script reports no running ECS task

Start the instance from its detail page, wait until its health is healthy,
and then rerun a newly copied script from the buyer portal.

The script reports an AWS access error

Use a buyer IAM user or role, not AWS root credentials. Confirm that the
one-time buyer shell-access policy is attached and use the latest script from
the instance detail page.

I need a different recovery point

Do not run an older script or modify the generated restore guidance. Request a
seller-admin restore and identify the desired backup date/time in the support
request.