The backup stories enterprise IT leaders actually read.
Volume 343, Issue 3 · 7 articles from 4 issues
Brien Posey explains how recoverability horizons, open data formats and regular restoration testing can help keep backups usable for years or decades.
Ensuring long-term data recoverability has become significantly more complex due to the shift toward cloud computing, Software-as-a-Service (SaaS) environments, and rapid software update cycles.
To illustrate these difficulties, consider attempting to restore legacy data from a SaaS application several years after the backup was created:
Overwriting Active Data: Restoring older data without unintentionally overwriting current live production data can be difficult. In many SaaS platforms, this requires setting up an entirely separate secondary tenant purely for legacy recovery.
Lack of System Consistency: Modern SaaS applications, underlying APIs, and backup software tools undergo constant evolution and frequent automated updates. Years after a backup is taken, the SaaS interface, database architecture, and local backup software will likely look completely different.
Irreversibility: Unlike legacy on-premises installations, organizations using SaaS applications cannot roll back the production application to a historical version to match the old backup schema.
Ranks nine enterprise backup platforms by coverage gaps, naming Eon best for cloud-first firms, Veeam for VMs and Commvault for SAP.
Analytics Insight cites a September 2026 census of 53,709 production cloud resources across 78 companies in which 61.2% had no recovery copy, with the median unprotected resource going 502 days without one.
It ranks nine backup platforms on how well they close coverage gaps, naming Eon best for cloud-first enterprises for its backup posture management and single-record recovery.
Veeam suits VM-heavy estates, Commvault SAP and Oracle, Rubrik cyber recovery, Cohesity legacy data centers, Druva endpoints and SaaS, and AWS Backup AWS-only teams.
Adastra's Obuks Akpede explains why green backup dashboards mislead and how to prove you can restore business services under pressure.
Immutable copies, defined RPOs and clean audit reports do not prove an organization can recover, says Obuks Akpede, a data governance consultant at Adastra.
In this Q&A he argues recovery must be designed into the operating model, with tested failure scenarios and cross-functional coordination on what comes back first.
Restores often fail on hidden dependencies such as DNS, identity, messaging, middleware and shared storage, and recovery data can itself become an attack path back in.
His top tip: stop tracking backup success rates and ask for evidence of the last full end-to-end recovery of the top business services.
The LTO Program now envisions 365TB LTO-14 cartridges, and the 40TB LTO-10 tape adds capacity by using longer, thinner tape.
HPE, IBM and Quantum, the companies behind LTO, now project LTO-14 cartridges with 365TB of native capacity, up from 40TB for the current extended LTO-10 media.
Capacity has grown 400-fold since LTO-1's 100GB in 2000 without changing the form factor, though the consortium has deliberately slowed the roadmap to prioritize reliability, cost per terabyte and manageable backup and restore times.
LTO-10 uses 15,104 data tracks, up 68.5% from LTO-9, and the 40TB cartridge uses a thinner Aramid base film to fit about 1,335 meters of tape.
LTO-10 dropped backward compatibility.
The October update brings Quick Machine Recovery to Windows Server 2025, so WinRE can pull fixes for mass boot failures.
Microsoft has added Quick Machine Recovery (QMR) to Windows Server 2025 through the October 2026 cumulative update, KB5126646.
When a faulty update, driver or configuration change stops servers from booting, QMR uses the Windows Recovery Environment to detect the failure, connect to the network and check Windows Update for a targeted fix, which it can download and apply automatically.
Administrators enable it through Group Policy and can set automatic retries.
Microsoft pitches it as a way to avoid fixing hundreds of servers by hand during large-scale outages.
Long-term backup survival depends on preserving the software, hardware, credentials, and documentation needed to restore.
Brien Posey recounts restoring decades-old media and finding backups he could read but not interpret, having lost track of the software, operating system, file format, compression, encryption, and passwords involved.
A backup must make sense to someone who did not create it, years later. He cites an insurance company that preserved its old tape drive, backup software, and a period PC, yet still lost data because it failed to keep the encryption key, an example of dependency erosion.
Restore dependencies should be treated as backup assets and managed with the same care as the backups themselves. Part 2 will cover why survivability has grown harder.
Organizations must ask if employees actually need their access levels to implement least privilege.
Least privilege best practices require asking whether employees genuinely need specific access, exemplified by a Portuguese hospital fined 400,000 euros for granting all staff doctor-level patient record access.
Implementation requires role-based administration, segregation of duties, and audit logging to ensure accountability. Organizations should inventory accounts, split backup roles, and establish ongoing management as staff moves between positions, treating this as an ongoing process rather than a one-time initiative.