Raiffeisen slashes mainframe cyber recovery to two hours

Raiffeisen Informatik has implemented a mainframe cyber recovery system with Hitachi Vantara that cuts recovery time in a total storage loss scenario from an estimated seven days to between one and two hours.

The Austrian banking IT provider runs infrastructure for institutions in the Raiffeisen Banking Group and for UNIQA. Its mainframe environment handles about one billion transactions a month for more than 1.25 million customers.

Previously, the group used two production storage systems across two data centre sites, supported by daily FlashCopies and batch backups written to a virtual tape library. That setup met a four-hour recovery target for standard failure scenarios, but the business identified a far greater risk when it modelled an attack that destroyed all storage systems across both sites at the same time.

In that scenario, recovery from conventional backup media was estimated to take about seven days. In a round-the-clock banking environment, restoring a consistent state is difficult because there is no simple point at which the entire landscape can be backed up together.

Instead, data would need to be rebuilt from individual tables and combined with forward-recovery procedures. For a core banking environment shared across several financial institutions, that process would be complex and difficult to complete within required service levels.

Air-gapped snapshots

To address the issue, Raiffeisen Informatik deployed Hitachi's Mainframe Cyber Resiliency Solution on its Virtual Storage Platform. The design combines continuous data replication with immutable snapshots, using two NVMe-based arrays as dedicated snapshot targets alongside existing storage systems at both mainframe sites.

All systems are linked by dark fibre. A key feature is the air gap between production storage and the snapshot targets, which means the production mainframe cannot address the protected snapshot storage.

According to the companies, an attacker with administrative access would therefore be unable to see, alter or delete the stored snapshots. New snapshots are created every two hours and remain write-protected during their retention period.

The architecture also allows recovery directly from a snapshot, rather than requiring all data to be copied back to the production storage environment first. That reduces the time needed to restore the full application estate to a consistent state.

Werner Georg Mayer, Managing Director at Raiffeisen Informatik, said the central requirement was the ability to restore the full environment in a coherent state.

"Without Hitachi Vantara's support, the solution simply wouldn't have worked. What matters is that from a snapshot, we can bring up the entire application environment with a consistent data state in one to two hours. A traditional file- or database-level restore of the full application landscape would take days - and achieving a consistent state across all systems that way is barely achievable," Mayer said.

Resilience testing

Raiffeisen Informatik tests recovery procedures every year in an isolated logical partition on the mainframe, segregated from the production network. The approach is designed to test response under severe disruption scenarios rather than rely only on documented recovery plans.

The work comes as financial institutions face rising scrutiny over operational resilience, including under the EU Digital Operational Resilience Act, which requires firms to maintain business continuity. The release cited research indicating companies lose an average of USD $300 million a year from unplanned outages, excluding potential regulatory penalties.

Octavian Tanase, Chief Product Officer at Hitachi Vantara, said the project focused on testing the most extreme failure case rather than a conventional disaster recovery model.

"Financial institutions need to know that their critical systems can be restored when conventional recovery assumptions fail. Raiffeisen Informatik has taken a rigorous approach to cyber resilience by designing and testing for a total-loss scenario and testing recovery procedures under realistic conditions. Reducing an estimated seven-day recovery to as little as two hours is significant, but the value goes beyond faster recovery. It gives mainframe leaders confidence that mission-critical banking services can be restored when it matters most," Tanase said.

The companies said the setup provides a recovery path for severe cyber disruption scenarios within Raiffeisen Informatik's contractual four-hour service level in the tested total-loss case. The two-hour snapshot interval also defines the maximum potential data loss in that scenario.

The project was completed in about four months from order to production by Raiffeisen Informatik's in-house team and Hitachi Vantara Professional Services. Raiffeisen Informatik has also signed an agreement to add VSP One Block High End to its mainframe storage environment.

"We have a clear view of what our mainframe environment needs to support the business going forward. Our storage infrastructure needs to keep pace with growing performance demands while giving us the flexibility to scale. VSP One Block High End gives us that foundation for the next phase of our mainframe strategy," Mayer said.