How To: Review Last Night's Backups
- Go to Dashboard. The seven-day rollup is on top with success rate, failures, and missed counts.
- Click Jobs and filter by the period you care about.
- Sort by status. Look first at Failed and Missed.
- Click a failed job to see the error message from the vendor.
- Click a missed device to see when it last succeeded and why the detector flagged it.
How To: Restore a File
- Go to Jobs, find the successful job that has the data, and click Restore From.
- Pick a restore type — file-level (one or more files) or an alternate location (a different device or path). There is no self-service full-device / bare-metal option, and the platform does not overwrite the live copy in place: recovered files land in a staging tree on the target device.
- Set the point-in-time snapshot you want.
- Enter the target device.
- Click Request. The restore is queued. Watch its status on the Restores page.
How To: Search Backup Contents for a File
- Go to Catalog.
- Enter part of the file path or name.
- Results show every backup version of matching files with size, checksum, and source device.
- Click a row and choose Restore to bring that version back.
How To: Investigate Coverage Gaps
- Go to Coverage → Gaps.
- Pick a window (30, 60, 90 days).
- Every device that has not had a successful backup in the window is listed.
- Click a device for its job history. Add a policy, fix the vendor configuration, or open a PSA ticket from here.
How To: Set an Alert Rule
- Go to Alerts.
- Click New Rule.
- Pick a rule type — failed job, missed backup, streak, size anomaly.
- Pick the severity, threshold, and notification channels.
- Tick Create Ticket if the rule should open a PSA ticket.
- Tick Trigger OnCall if the rule should page the on-call rotation.
- Save.
How To: Produce an SLA Report
- Go to Reports → SLA.
- Pick the period and the clients you want to include.
- Click Calculate. The live calculation runs and returns the SLA percentages.
- Click Save to persist the report to the Reports list. Saved reports are available for download later.
How To: Cancel a Running Backup Job
- Open the job from the Jobs list.
- Click Cancel. The cancellation request is sent to the vendor.
- Status moves to Cancelled once the vendor confirms.
How To: Set Up a New MSP Client in Backups
- Go to Clients and click New Client.
- Enter the name, domain, primary contact, and PSA company id.
- Save. Once the vendor path is collecting, jobs from connected vendors that match the company mapping land under this client. ⚠️ That path is built but not yet collecting — no tenant has a vendor configured and end-to-end collection has not been proven — so no vendor-fed jobs land here today.
Frequently Asked Questions
What vendors are supported? Veeam, Datto, Acronis, Cove, Backup Exec, and others — each added through a normalized connector and surfaced through the same dashboard. ⚠️ Built but not yet collecting: no tenant has a vendor configured and end-to-end collection has not been proven, so this path shows nothing today.
What counts as "missed"? A device has not recorded a successful job within its expected interval. The check runs every fifteen minutes and creates a synthetic missed job when it fires.
When does the system page OnCall? When a device has three or more consecutive failed jobs in 48 hours, or when a missed backup exceeds 3x the expected window, OnCall is triggered.
Where are vendor credentials stored? In Azure Key Vault, referenced by a secret name on the vendor record. They are never stored in the database in plaintext.
Can I pull ad-hoc job data instead of waiting for the sync? Yes. Open the vendor and click Sync Now.
Does the catalog contain the actual file contents? No. It contains metadata — path, size, checksum, version. The bytes are pulled from the platform's own object storage when you run the restore.
Will an unsupported vendor show up? Only vendors with a connector implementation feed jobs into Backups. For others you can still record manual job status through the ingest API.
How do I know the compliance report is accurate? It is calculated from the same job records you see in the Jobs list, filtered by the policies you set. You can export the underlying rows.
Troubleshooting
Vendor sync is failing. Open the vendor. The Last Sync Error field shows the reason. Most commonly: expired API key, revoked permission, or vendor API outage. Rotate credentials via Key Vault and click Sync Now.
Job shows as Missed but it did run. Check that the vendor's job is mapped to the right device and client. If the external_id shape changed, the upsert may have created a new record. Merge manually or re-map the company.
Restore request stuck at Pending. An endpoint restore is claimed by the device's agent on its next poll. If the request stays at Pending, confirm the target device is online and its agent is running — the request is served on the next poll.
PSA ticket was not created on a failed job. Confirm the alert rule for that rule type has Create Ticket enabled and that PSA integration is configured in Settings. The ticket id (or a fallback id) is saved on the job so you can trace it.
OnCall page did not fire. Confirm the rule has Trigger OnCall enabled, OnCall integration is configured, and the on-call schedule has an active responder.
Coverage gap shows a device I retired. Mark the device as Deleted in the source-of-truth (typically CMDB or RMM) or hide it from reporting by removing it from the client's active device list.
Storage utilization is wrong. Storage numbers come from the org's storage settings and the platform's own captured usage; usage reported by a connected vendor is included when that path is live. Click Recalculate on the storage page, or update the quota value if it has changed.
Cron jobs are not running.
The embedded scheduler may be disabled. Check with an administrator — if BACKUPS_CRON_ENABLED is set to false, missed-backup and streak detection do not run.