AWS Partner Network (APN) Blog
Accelerate Your Ansys Simulations with Ansys Gateway Powered by AWS
By Rob Markworth, Sr. Principal R&D Engineer – Ansys
By Dnyanesh Digraskar, Sr. Partner Solutions Architect – AWS
By Tomo Sakatoku, Principal Partner Solutions Architect – AWS
By Raymond Chow, Sr. Partner Solutions Architect – AWS
 |
| Ansys |
 |
To meet challenging product design goals and deadlines, today’s complex engineering projects need instant access to state-of-the-art computing infrastructure as well as advanced computer-aided engineering (CAE) and computer-aided design (CAD) software.
However, procurement and configuration of these resources at scale can involve large up-front costs and long deployment timelines. Ansys and Amazon Web Services (AWS) announced a strategic collaboration in 2022 that introduced Ansys Gateway powered by AWS, which facilitates seamless access and deployment of Ansys products on the AWS Cloud.
Hosted in AWS Marketplace, Ansys Gateway is a solution for developers, designers, and engineers looking to manage their complete Ansys simulation workflows in their own AWS cloud environment. Users can run any of the 50-plus Ansys products on virtual desktops or high-performance computing (HPC) clusters through a single, unified interface.
In this post, we cover the architecture details of Ansys Gateway and explain the principal architectural components along with a typical customer workflow for using Ansys Gateway.
Solution Architecture
The architecture for Ansys Gateway is developed based on the AWS Well-Architected Framework best practices focusing on operational excellence, security, reliability, performance, cost, and sustainability.
Figure 1 – Architecture diagram of Ansys Gateway powered by AWS.
The Ansys Gateway solution is made up of two primary components—Control Plane and Application Plane. The Control Plane refers to the services deployed and hosted in Ansys’s AWS account. The Application Plane refers to the services and infrastructure that is orchestrated in a customer’s own AWS account.
For more details on the architecture concepts, refer to the AWS whitepaper on SaaS architecture fundamentals.
Control Plane
The Control Plane consists of the following principal components, as noted in Figure 1 above:
- Web Service: Provides the user interface (UI) to allow Ansys Gateway customers to follow guided-access wizards to create, view, and access HPC clusters, virtual desktop infrastructure (VDI) instances, or general supporting resources. It also provides the interface to manage access, usage reports, budget and security control, and general Ansys Gateway administration. These are based on automatically scalable Amazon Elastic Beanstalk instances that scale to meet current demand.
- API Service: The core business logic where, with permission from the customer, the Amazon Elastic Compute Cloud (Amazon EC2) resources are requested, configured, and maintained. This set of servers handles all functions; from receiving user interface instructions and reporting on status, to deploying Ansys simulation software and managing user access.
- Amazon CloudWatch: The servers are a critical component of the system that’s supported with Amazon CloudWatch for up-to-date monitoring and reporting to ensure operational efficiency and resilience. The server elements are stateless, so all necessary data is stored in a resilient Amazon Relational Database Service (Amazon RDS) database. Meanwhile, the Amazon Simple Storage Service (Amazon S3) package store contains cloud-optimized packages to automatically install and configure user-selected Ansys applications, including the automated setup of HPC clusters.
- AD Connector: Allows Ansys Gateway to join the Amazon EC2 resources to the customer’s Active Directory.
- Communications Proxy: Allows Ansys Gateway to perform and manage communication between various web resources.
- Compute Agents: These agents are installed on the EC2 instances to provide the management capability for Ansys Gateway to install, configure, and manage each instance.
Application Plane
The Ansys Gateway Application Plane consists of elements that are provisioned by users in their own AWS accounts for running Ansys simulation workloads. The following figure represents a typical architecture of the Application Plane resources.
Figure 2 – Data plane and provisioned resources in user’s AWS accounts.
Customer Onboarding
Starting with a secure onboarding experience, the customer is directed from AWS Marketplace to the Ansys Gateway onboarding wizard. This is a guided entry point to provide details, set up prerequisites, and commission the AD Connector.
Access to the Control Plane is controlled by the customer’s identity provider (IdP). Active Directory (using the Gateway AD Connector) and AWS IAM Identity Center (successor to AWS Sign-On) using SAML are supported to allow a user to log in.
Within the customer’s Active Directory, security groups are created to allow the customer to specify users and administrators. When accessing virtual machines or clusters, customers can log in with their domain credentials or via local administrator accounts created by Ansys Gateway.
Project Space
Ansys Gateway allows users to configure governance and usage-related controls for the AWS resources provisioned in their accounts. A dedicated project space can be created for executing workflows on a per-user or department basis.
Figure 3 – Project spaces as configured by users.
For each project space, users can set budgets for resources related to a single workflow or department, and manage role-based access. Resources such as license servers or data stores can be easily shared across project spaces to allow efficient operation of multiple types of workflows.
Simulation Workflow: VDI and HPC Cluster Creation
Using a guided step-by-step process, customers can create either VDI or cloud-optimized HPC clusters in a few steps. The Ansys Gateway interface provides recommended options for operating system, Ansys application versions, EC2 instance types, and disk options used to create a VDI or HPC cluster.
Users can access recommendations for specific software and hardware configurations that have been tested and optimized by Ansys. These recommendations provide a quick start with common configurations, while also allowing access to the full range of EC2 options to meet custom workload requirements.
Simulation Workflow: License Connectivity
Ansys Gateway leverages a customer’s existing Ansys license pool. Those licenses might be managed on an existing license server in the customer’s network.
A new license server can also be configured and deployed by the customers directly on AWS. Once defined, the license server package can be shared across multiple project spaces and users.
Simulation Workflow: Running Ansys Solvers
The solver settings and workflows in Ansys Gateway are tailored to each specific Ansys application. For HPC, the job submission steps follow best practices for the specific Ansys application being used.
The steps below are generally applicable to all applications, and the details are well documented for each application. You can see detailed documentation for running solvers using Ansys Gateway.
An interactive virtual desktop mode (VDI mode) is available for users running Ansys applications on a single Amazon EC2 instance, or for accessing the head node of a cluster. In interactive mode, a user can perform the following steps:
- Preprocess a model or generate a mesh on a high memory instance.
- Submit an HPC job to a cluster and monitor the run.
- Post-process or visualize the results on powerful graphics-accelerated cloud hardware.
Resources in Ansys Gateway are intended to be dynamic and under the user’s control. The user can stop VMs and cluster nodes when they’re not needed to minimize costs. A stopped resource maintains its configuration and storage, and can be converted in to the “Running” state on demand.
Along with manually stopping the instances, the Ansys Gateway user interface provides the ability to set up instance-state timers depending on key metrics such as CPU utilization, current logged-users users, or memory utilization. Users can customize these timers to ensure the instances are not running when idle, to avoid paying for idle resources.
Depending on the memory or compute requirements for running an Ansys application, the user can choose an appropriate Amazon EC2 instance type before starting the VM. The list of recommended EC2 instances for running various Ansys applications is included in the Ansys Gateway documentation.
To resize an HPC cluster, the user can increase or decrease the number of nodes directly through the Ansys Gateway UI to meet the needs of the next HPC job.
Simulation Data Management
The data used in Ansys Gateway correspond to the native file types associated with an Ansys application. Simulation inputs such as geometry file, mesh, scripts, subroutines, and macros can be created in-situ during a VDI session, or can be copied from a shared repository or other AWS native storage options.
Simulation outputs consisting of result files, log files, and custom outputs are typically large in size so it’s best to post-process them on a VDI session. As needed, files or subsets of files can be transferred to permanent storage or a user’s desktop. Refer to the Ansys Gateway best practices for setting up storage.
Data Security
Ansys Gateway secures each data location with AWS Access Control List (ACL) controls. All customer data is generated and stored in the customer’s own AWS environment, and is secured with encrypted data connections. The customer’s data never leaves the customer’s network.
Summary
Following are the benefits of running your Ansys workloads on Ansys Gateway powered by AWS:
- Bring your own Ansys license to install and run Ansys simulation products on AWS infrastructure.
- Manage and control your CAD/CAE cloud consumption directly within your AWS subscription.
- Increase pace of innovation with virtually unlimited and flexible AWS compute capacity.
- Instantly access the latest technology for compute, storage, and networking.
- Leverage AWS global presence to enable efficient collaboration and management of engineering simulation resources.
Ansys Gateway is a solution for developers, designers, and engineers who want to manage their complete Ansys simulation workflows on AWS. Engineers can access cloud computing resources from anywhere on virtually any device via web browser.
Get a free trial of Ansys Gateway powered by AWS on AWS Marketplace.
.
.
Ansys – AWS Partner Spotlight
Ansys is an AWS Partner that empowers customers to design and deliver transformational products that push boundaries with the predictive power of simulation.