Dr.Karl-Remeis-Sternwarte Bamberg - Astronomisches Institut

Prof. Wilms's research team is focusing on X-ray Astronomy and is also involved in the development of several new X-ray telescopes to be launched in the next decade. They use software called SIXTE, which helps them simulate data expected from the telescope. Currently, the software runs on university servers and needs about 512GB of memory. However, they are facing challenges in scaling the systems and are considering cloud alternatives to meet their simulation needs.

SIXTE software simulates X-ray instruments on scientific satellites for missions usually led by NASA and ESA. It's also used for instruments like eROSITA on the Russian-German Spectrum-X-Gamma mission and for Chinese Academy of Sciences projects. The software predicts how observations of celestial objects would look with a given satellite design.

It's modular, written in C++, and available under the GPL. SIXTE includes sensor physics, X-ray imaging modules and tools for building data analysis pipelines. Users include scientists planning science cases, instrument developers evaluating performance, and software developers working on the satellite mission's ground segment.

Goal

‍The research group aimed to implement a scalable solution for hosting and running their SIXTE simulation software on cloud infrastructure, ensuring redundancy, scalability, and durability. This transition from on-prem servers to the cloud required clarity on the cost, performance, and usage of cloud systems. To achieve this, Ankercloud provided a Proof-of-Concept setup, designing, setting up, hosting and training the research group on using AWS servers. Ankercloud also conducted benchmarking tests to compare price-performance benefits across multiple AWS instance types and recommended the most suitable setup for the team on AWS.

The solution aimed at creating a comprehensive cloud infrastructure to accommodate the SIXTE simulation software needs of the research group effectively. It encompassed:

1. VPC Configuration: Established a dedicated Virtual Private Cloud (VPC) in the us-west-1 (Ohio) region, covering all Availability Zones (AZs). Linked the VPC with essential components like DHCP option sets, main route tables, and Network ACLs and configured 4 subnets within, including 2 public subnets for internet access and 2 private subnets for internal resources.
2. Internet Access: Implemented an Internet gateway for internet traffic within public subnets, with configured route tables. For private subnets, set up secondary route tables directing outbound traffic to a NAT Gateway.
3. Security Measures: Created security groups to regulate traffic for EC2 instances, ensuring secure communication within the infrastructure by adding necessary rules.
4. EC2 Instance Deployment: Launched EC2 instances using CloudFormation stack, focusing on suitable instance types like r5a.8xlarge. Attached these instances to designated security groups for access control enforcement.
5. Application Setup: Installed the SIXTE software on EC2 instances by downloading and installing zipped files of SIXTE & SIMPUT software, followed by compilation and installation of the simulation software.
6. AMI Creation: Generated Amazon Machine Images (AMIs) for EC2 instances to enable replication and deployment. Ensured software functionality verification and provided clear identification through naming and description.
7. Cloudformation Template: The solution utilized CloudFormation templates stored in an S3 bucket to efficiently create the cloud infrastructure for hosting the SIXTE simulation software. These templates automated the provisioning process, ensuring consistency across environments. With CloudFormation, the solution provided scalability and flexibility, allowing the research group to easily adapt to changing demands and conduct simulations effectively in a reliable cloud environment.
8. HPC and Parallel Clusters: Setting up an HPC (High-Performance Computing) and parallel cluster on AWS infrastructure offers a scalable, cost-effective, and efficient solution for transitioning the SIXTE simulation software to the cloud for X-Ray Astronomy research. This setup facilitates scalability by distributing simulation workloads across multiple computing nodes, ensuring timely completion. Redundancy measures like backup nodes and data replication ensure uninterrupted operation. Benchmarking tests conducted by Ankercloud help optimize performance and control costs, while centralized management tools streamline resource allocation and job scheduling, empowering the team to focus on their research with confidence in their cloud infrastructure.

The business outcomes of implementing this infrastructure for hosting the SIXTE simulation software include:

1. Scalability and Flexibility: The setup allows the research group to scale their simulation capabilities based on demand, without the limitations of on-premises servers. This flexibility enables them to adapt to changing research needs and workload requirements.
2. Cost Optimization: By leveraging cloud services, the research group can optimize costs by paying only for the resources they use, avoiding upfront hardware investments and reducing operational expenses associated with maintenance and upgrades.
3. Improved Accessibility and Collaboration: Hosting the software on the cloud enhances accessibility for researchers, allowing them to access the simulation environment from anywhere with an internet connection. This facilitates collaboration among team members and potentially with external partners or collaborators.
4. Enhanced Security and Reliability: The use of AWS services provides robust security features and reliability, ensuring the integrity and availability of the simulation environment. This helps protect valuable research data and ensures continuity of operations even in the event of hardware failures or disruptions.
5. Faster Time-to-Insight: With the ability to quickly provision and scale resources as needed, researchers can accelerate the simulation process and obtain insights faster. This agility enables them to iterate more efficiently on their research hypotheses and experiments, potentially leading to groundbreaking discoveries in astrophysics.

Summary, transitioning to cloud infrastructure for hosting the SIXTE simulation software offers numerous advantages to the research group. It provides scalability and adaptability, enabling them to respond to changing research demands without the limitations of traditional servers. Cost efficiencies are realized by paying only for utilized resources, eliminating upfront hardware investments and reducing ongoing operational costs. Enhanced accessibility and collaboration foster seamless teamwork among researchers and external partners, facilitating knowledge sharing and innovation. Agile resource provisioning accelerates the simulation process, leading to quicker insights and potentially groundbreaking discoveries in astrophysics. Overall, the adoption of cloud infrastructure empowers the research group to propel their scientific pursuits with greater efficiency and effectiveness.

About Ankercloud

Ankercloud has been delivering projects utilizing AWS services & helping clients across the globe. Ankercloud is an advanced partner for AWS with a long experience of delivering solutions on multi-cloud platforms. A team of professionals closely discuss the requirements & architect the solution approach in an agile model of project delivery. A hands-on consulting approach is encouraged at all levels within the organization to gain the customers’ trust and respect. This provides Ankercloud with a competitive edge in the partner ecosystem as a preferred choice.

About Amazon Web Services

‍Amazon Web Services (AWS) has helped thousands of organizations migrate to the cloud, from global enterprises to small businesses. The customer chose AWS as it has significantly more services and more features within those services, than any other cloud provider – from infrastructure technologies like computingcompute, storage, and databases,  – to emerging technologies, such as machine learning and artificial intelligence, data lakes and analytics, and Internet of Things. It was the best choice for them to be future-ready.