Confined to Land, Data Could Not Cross Borders
A Singapore-based contractor responsible for the design and maintenance of cross-border oil and gas pipeline infrastructure faced a regulatory and operational impasse. Engineering teams were running complex fluid dynamics simulations and pipeline stress analyses using proprietary software licensed per seat and tied to physical workstations. All project data, including geological surveys and integrity reports, was classified under national energy security frameworks, prohibiting storage or processing in foreign data centres. With project timelines tightening and collaboration demands increasing between on-site engineers and design offices, the existing setup—relying on isolated high-performance desktops and USB-based file transfers—was no longer tenable. The inability to scale compute resources on demand threatened delays on a USD 42 million pipeline integrity upgrade, with penalties accruing for every day past deadline. The company needed a way to virtualise compute-intensive workloads without compromising data sovereignty or violating licensing agreements.
Legacy Infrastructure Was Never Designed for Virtual Workloads
Upon audit, we found the client’s IT environment consisted of standalone workstations running Windows and Linux in parallel, each with locally attached storage and no centralised backup. Software licenses for simulation tools from ANSYS and Schlumberger were node-locked, preventing redistribution. No virtualisation layer existed, and the network lacked segmentation or monitoring. Engineers manually copied multi-gigabyte simulation files between locations using encrypted USB drives, introducing version control issues and data leakage risks. Worse, no disaster recovery plan was in place—any hardware failure could result in days of downtime. The client assumed that because public cloud was off-limits, virtualisation was not an option. They were unaware that a properly architected on-premise private cloud could meet both regulatory and performance requirements.
We Built a Virtualisation Platform Anchored in Local Control and Performance
We deployed a VMware vSphere-based private cloud on HPE ProLiant servers with redundant power and fibre-channel storage from Pure Storage, installed within the client’s secured Singapore data room. The infrastructure was segmented into dedicated clusters: one for simulation workloads, another for data storage and backup, and a third for management and monitoring. Each ANSYS and Petrel license was migrated to a virtual machine with GPU passthrough enabled, ensuring compliance with vendor terms while allowing engineers to access their environments remotely via secure PCoIP connections. We integrated Veeam Backup & Replication to enable hourly application-aware backups with immutable storage, meeting internal audit requirements. Network traffic between virtual machines and storage was encrypted in transit using FIPS 140-2 compliant protocols, and access was enforced through role-based controls integrated with the client’s existing Active Directory. The entire stack was designed for non-disruptive maintenance, with live migration and fault tolerance enabled for critical VMs.
Simulation Cycles Accelerated, Downtime Reduced to Minutes
After deployment, engineers reported a 60% reduction in simulation job wait times due to dynamic resource allocation across the cluster. The ability to spin up additional VMs during peak analysis periods eliminated bottlenecks that previously delayed reporting by days. Uptime for critical engineering applications increased from 92% to 99.95%, with only 22 minutes of unplanned downtime recorded over a 12-month period—mostly during scheduled maintenance windows. Data recovery time dropped from an average of 14 hours using physical backups to under 12 minutes using Veeam’s instant VM recovery. The client completed the pipeline integrity project two weeks ahead of schedule, avoiding contractual penalties and securing a follow-on contract for monitoring system integration. The private cloud now supports three additional projects, with all data remaining within Singapore’s jurisdiction.
Regulatory Constraints Can Drive Better Infrastructure Decisions
For Singapore-based firms handling sensitive technical data, especially in energy, transport, or defence sectors, data residency rules are often seen as obstacles. But when treated as design requirements rather than limitations, they can lead to more secure, efficient, and scalable architectures. This client’s compliance needs forced a re-evaluation of infrastructure strategy, resulting in a system that not only met regulatory standards but outperformed their legacy setup in every operational metric. The lesson isn’t to avoid the cloud—it’s to choose the right cloud, even if it means keeping it on the same floor as the engineers who use it.
For firms managing sensitive engineering workflows under strict data control mandates, exploring on-premise virtualisation with TYPENT can turn compliance into a competitive advantage.