Esteyco expands geographically separated storage infrastructure

Esteyco expands geographically separated storage infrastructure

Spanish engineering consultancy Esteyco is extending its storage infrastructure across two locations more than 500km apart, introducing asynchronous replication and immutable snapshots to protect demanding research workloads.

Esteyco expands geographically separated storage infrastructure
Summary
  • Esteyco has deployed Hitachi VSP One Block to support high-performance engineering simulations, including work associated with ITER.
  • The company is establishing a second system in Barcelona, more than 500km from its primary installation, with asynchronous replication.
  • Supplier-reported testing reduced a storage maintenance window from several hours to approximately five minutes, while immutable snapshots and additional backups support recovery.

Spanish engineering consultancy Esteyco is extending its data infrastructure across two geographically separated locations, introducing storage replication between a primary installation and a second system being established at its Barcelona data centre.

The project, delivered with Hitachi Vantara and Quadrics Sistemas, uses Hitachi Virtual Storage Platform One Block equipment to support demanding engineering simulations and protect research data against interruption or loss.

The new arrangement was detailed on 18 November. Esteyco is establishing a second storage system more than 500km from the primary site, with asynchronous replication intended to maintain an additional copy of critical information.

The installation supports the company’s work on complex engineering projects, including simulations associated with ITER, the international nuclear fusion research programme.

Esteyco’s activities involve structural analysis, modelling of critical components and other computing workloads that can run for extended periods. Interruptions affecting the underlying storage infrastructure can therefore have consequences beyond the immediate availability of individual files.

According to the companies involved, previous storage performance and stability problems could require engineers to restart long-running simulations, creating additional demands on computing resources and project schedules.

Quadrics evaluated the new equipment using production workloads before proceeding with implementation. The project included virtualising the existing storage environment through VSP One Block.

Supplier-reported results indicate that the required maintenance window was reduced from several hours to approximately five minutes. The figure describes the reported maintenance operation rather than a guaranteed recovery time following every possible failure.

Storage availability influences computing performance

High-performance engineering workloads depend on access to large datasets, intermediate simulation results and files created during calculations.

Some workloads involve intensive writing to storage, particularly when applications save intermediate states or produce substantial volumes of output for subsequent analysis.

Where storage performance becomes inadequate, computing systems may spend time waiting for information to be written or retrieved, reducing the effective use of installed processing resources.

Stability problems can be more disruptive. If a long-running calculation cannot continue after a storage interruption, engineers may need to repeat work that has already consumed substantial computing time.

Quadrics reported that Esteyco’s workloads included periods of intense write activity, creating demanding conditions for the storage equipment.

Testing the proposed infrastructure with production workloads allowed the engineering team to assess its behaviour under operating conditions relevant to the organisation.

Virtualising the existing storage environment can also affect maintenance arrangements by changing how underlying hardware is presented to applications and how data is moved between systems.

Esteyco’s reported reduction in maintenance time suggests that the new arrangement has improved a specific operational procedure, although the public project information does not establish comparable results across every type of maintenance or failure.

Storage performance alone does not determine the availability of an engineering computing environment. Processing equipment, networking, software and the facility’s electrical and cooling systems must also remain operational.

The storage upgrade addresses one part of that dependency, particularly the reliability and accessibility of information used by long-running simulations.

Replication and backups address different risks

Esteyco’s second VSP One Block installation is being established at its Barcelona data centre, more than 500km from the primary system.

The two locations will use Hitachi Universal Replicator to transfer information asynchronously, providing an additional geographical copy without requiring every write operation to be completed simultaneously at both facilities.

Asynchronous replication can reduce the performance impact of maintaining remote copies, particularly where the distance between sites introduces network latency.

However, replication that occurs after the original write can leave a time difference between the most recent information at the primary site and the copy held elsewhere.

The potential amount of information affected by a failure depends on the replication configuration, network conditions and how far the secondary copy is behind the primary system.

Geographical separation can reduce exposure to incidents confined to one site, such as a local utility interruption or infrastructure failure.

It does not remove dependencies involving shared software, configuration errors, compromised credentials or common network services.

For that reason, Esteyco’s arrangement also includes immutable snapshots intended to support recovery from data corruption or malicious changes.

Immutable snapshots provide protected recovery points under the applicable storage configuration, but their usefulness depends on retention settings, access controls and whether suitable recovery procedures have been established.

Quadrics is additionally supporting data retention through secondary storage, object storage synchronisation involving a managed service provider, and regular tape backups.

These arrangements provide different recovery options rather than relying exclusively on replication between the two primary storage systems.

The distinctions are important because corruption or unintended deletion can potentially be transferred between replicated environments. Independently maintained recovery copies can provide additional protection where configured appropriately.

The published project information does not specify storage capacity, replication bandwidth or contractual recovery objectives, limiting direct comparison with other geographically separated installations.

Nor does a second storage system establish that Esteyco has created a fully operational alternative computing site. Applications, processing resources and network access would need to be available for complete workload recovery.

The project nevertheless represents a physical infrastructure investment spanning separate facilities, with changes to maintenance, storage architecture and data protection for computationally intensive engineering work.

Its operational performance will depend on successful commissioning of the Barcelona installation and continued testing of replication, recovery and maintenance procedures across the two locations.


Stay updated with the latest insights and trends in the data centre industry by subscribing to our newsletter.

← Back

Thank you for your response. ✨