Implementing AWS end user computing for Oil Search
Implementing Amazon end user computing services to provide high performance computing and a familiar and consistent desktop experience across geographies
- Secure, scalable defence-in-depth and compliant AWS foundation
- Scalable VDI solution to support graphics intensive applications
- Access to high performance computing without location and physical hardware restraints
Papuan Oil Search Limited (Oil Search) is a Papua New Guinea (PNG) and Australian listed company with regionally significant investments in oil and liquefied natural gas (LNG) and an extensive appraisal and exploration portfolio in PNG.
With an expanding presence across multiple geographies and with assets in remote locations where access to technology can be limited, Oil Search operates in a challenging IT environment.
A complicating feature of the Oil Search IT operation is that it is hosted in Data Centres in each geography, which are not replicated across each other.
A key application used by Oil Search engineers and geophysicists is Petrel. Petrel is a modelling, mapping and correlating software platform used in exploration and production that needs specific high performance computing hardware to operate.
Oil Search provided compute capacity for Petrel through high-cost physical desktops that included HPC hardware, thus limiting the flexibility of user location.
The large Petrel data sets also presented a challenge as they needed to be located close to the application to reduce latency and improve its performance.
Looking for ways to provide access to Petrel in a more efficient and scalable way, Oil Search trialled the use of Amazon AppStream 2.0, undertaking a Proof of Concept for the Petrel application. The PoC identified AppStream as a likely alternative offering substantial operational improvements for users of Petrel.
CMD Solutions was engaged to progress the introduction of AppStream 2.0 at Oil Search – with complementary Amazon Workspaces services – from PoC to a production deployment for Petrel.
The initial EUC project was constructed in the form of a pilot engagement. Success criteria were formed during the initial workshops and an agreed work backlog was established.
After capturing Oil Search’s environment requirements an appropriate AWS foundation was agreed and created. It was built using a multi-account defence-in-depth design.
The AWS accounts included tiered subnets for management VPC and separated subnet tiers for VDI Clusters, security groups and Network Access Control Lists (NACLs) to secure network traffic routing.
Amazon AppStream 2.0 and Amazon Workspaces were used to address Oil Search’s need to provide end user compute services to users at a reduced TCO with a low management overhead.
Strategies to deliver these efficiencies included scheduling AppStream 2.0 cluster to cyclically scale up in business days/hours to the desired running capacity and scale down out of business hours to a minimum pilot light infrastructure, and configuring ALWAYS ON vs On DEMAND Amazon Workspaces on a user by user basis, based on usage profile.
A single Identity Provider and Azure Enterprise application were created to allow access per environment between Oil Search and AWS for the Petrel Stack.
The use of AWS Lambda and Active Directory (AD) enabled the provisioning and decommissioning of Amazon Workspaces automatically without manual intervention, gaining the security benefits of each user authenticating to AD and enforcing the Group Policy configuration.
From an inflexible and infrastructure based environment, Oil Search have emerged with a modernised AWS foundation landing zone that utilises Amazon AppStream 2.0 to give access to a core business app – Petrel – in more locations and with high performance computing capability without a desktop computer set-up.
Through Amazon Workspaces, Oil Search users can now have a familiar and consistent desktop experience across geographies, with Workspaces doing the heavy lifting to replicate data centres across the Oil Search IT environment.
Desktops can be managed effectively according to demand across the business, and processing scaled accordingly.
The CMD team undertook a complete training, handover and solution delivery to the Oil Search IT team in an effort to ensure its AWS foundation and services continue to provide ongoing and expanding value to the Oil Search business.