“Multi-tier backup” refers to a backup strategy in which backup copies are stored across multiple storage tiers and - typically automatically - moved or additionally replicated between those tiers. Common tiers include fast local disk (for quick restores), cost-efficient cloud object storage (for longer retention), and tape or cloud archive storage (for long-term retention and compliance). The goal is to balance recovery objectives (RTO/RPO), cost efficiency, and retention requirements.
Policy-based tiering: Rules that define when and based on which criteria (age, frequency, data class) backups are written to or moved into a specific tier.
Automated lifecycle & retention management: Retention schedules, versioning, and automated expiration/deletion in line with defined policies (e.g., 30/90/365 days).
Support for multiple backup targets: Backup to disk, dedupe appliances, NAS/SAN, object storage (e.g., S3-compatible), cloud archive, and tape—staged or in parallel.
Copy & replication orchestration: Creating additional copies to other tiers/sites (e.g., offsite/cloud), including time-window and priority control.
Deduplication & compression: Reducing storage consumption, especially for mid-cost and low-cost tiers.
Encryption & key management: Protecting data at rest and in transit, optionally with centralized key management.
Immutability / WORM & ransomware protection: Time-locked, tamper-resistant backups, air-gap options, and safeguards against manipulation.
Backup verification & restore testing: Integrity checks, automated verification, and scheduled test restores.
Tier-aware restore optimization: Features such as staging (rehydrating from archive to a fast tier), granular recovery, and prioritization of critical systems.
Bandwidth & cost control: Throttling, cloud transfer schedules, WAN dedupe, and reporting on storage and transfer costs.
Central monitoring, reporting & auditing: Visibility into backup health, SLA compliance, per-tier capacity, and compliance evidence.
A company keeps daily backups for 7 days on local disk for fast recovery, then moves them to cloud object storage for 90 days.
Weekly full backups are additionally replicated to a secondary data center, while monthly backups are automatically offloaded to an archive tier (e.g., cloud archive or tape) for several years.
For mission-critical data, two parallel copies are created: one on a dedupe appliance (fast restore) and one immutable copy in object storage (ransomware protection).
A “disk-to-disk-to-tape” (D2D2T) approach: backups land on disk for operational restores and are later exported to tape for long-term retention and an offline air gap.
A SaaS provider uses short-lived snapshots on high-performance storage for very low RTOs and later moves consolidated backups to lower-cost tiers to reduce expenses.