In the realm of High-Performance Computing (HPC) and Data Analytics, remote desktop solutions are essential for accessing and managing intricate computing infrastructures. ThinLinc from Cendio has emerged as a transformative tool in this space, offering significant advantages for users interacting with sophisticated computing environments. This article explores how ThinLinc addresses complex computing challenges and highlights spinTwo’s role in implementing and enhancing ThinLinc’s capabilities.
Expertise in Complex Computing Solutions
spinTwo, based in the U.S., differentiates itself with a rare combination of scientific insight and technical proficiency. Unlike traditional IT service providers, spinTwo’s team includes former scientists who deeply understand the computing requirements of the scientific community. This background enables spinTwo to tailor HPC, Kubernetes, and AI-focused solutions precisely to meet scientific workflows, optimizing performance to support research needs. Their dual expertise facilitates seamless collaboration between IT and research teams, ensuring that technical implementations align closely with scientific objectives.
But What is ThinLinc?
ThinLinc is a powerful remote desktop solution designed to provide secure, high-performance access to Linux-based systems. Developed by Cendio, ThinLinc enables users to interact with remote computing environments, including complex HPC (High-Performance Computing) setups, with a graphical interface. It supports advanced features like GPU acceleration and multi-user access, making it an ideal tool for managing and utilizing sophisticated computing resources from virtually any location.
Cendio has identified important metrics based on the user experience of using ThinLinc read this article:
User Satisfaction and Importance
- 81% of respondents reported that they were satisfied or extremely satisfied with the services.
- 75% of respondents reported that the remote desktop services are extremely or moderately important to their ability to use HPC resources.
Growth in User Adoption
In 2014, Indiana University (IU) launched Karst Desktop (KD), a VNC-based desktop frontend for Karst, its Linux-based cluster, based on the ThinLinc product by Cendio. By the end of 2018, more than 1,000 (one thousand) users had used KD, and more than 900 of those had used the service in that year. The KD service has helped increase the total number of HPC users at IU by more than 50%.
Usage Statistics at Indiana University and Purdue University
- At Indiana University, desktop services are used by roughly 65% of all HPC users, and among all the HPC systems, RED (Research Desktop), has the most users.
- Purdue University also averages 300+ users every month.
Perceived Benefits
Several items asked respondents to indicate the importance of various features of the desktop interface. The ability to run long uninterrupted sessions was rated most important by both university populations. Applications with graphical interfaces can be offered; links to server-based applications such as Galaxy can be provided as links on the desktop; and terminals are available to run traditional command-line based software.
Uninterrupted Sessions
This feature is reported as one with the highest rating of importance when using ThinLinc.
Enhancing Remote Desktop Access
In the context of how ThinLinc could be integrated into the HPC software stack, it’s important to highlight its role in enhancing the overall computing environment. ThinLinc’s architecture, which includes its server, client, and connectivity with HPC systems, plays a critical role in optimizing remote desktop access. This integration is particularly beneficial for environments that require high-performance graphical applications, as ThinLinc supports advanced features like GPU acceleration and secure connections.
By efficiently managing remote desktop access, ThinLinc becomes an indispensable tool for system administrators and users within an HPC environment. The process—from user authentication to accessing and interacting with graphical applications on HPC resources—illustrates how ThinLinc fits seamlessly into the HPC software stack. This capability not only enhances user experience but also ensures that the HPC infrastructure is used to its full potential, making ThinLinc a valuable addition to any advanced computing setup.
ThinLinc for System Administration
For many traditional system administrators, tools like screen sessions have long been sufficient for managing remote Linux machines, whether they are standard or critical systems. However, ThinLinc goes beyond these basic tools by offering not only command-line access but also a user-friendly interface for graphical applications. This is particularly valuable in environments where users need to interact with graphical software remotely, as ThinLinc’s intuitive interface makes it accessible even for those less familiar with command-line tools.
A key feature of ThinLinc is its support for multi-user access, a major benefit of UNIX-based systems. It allows multiple users to run concurrent sessions on the same server, each with a complete graphical environment, enabling effective resource sharing while maintaining individual working environments. Moreover, ThinLinc’s multi-session capability lets users suspend a session and resume it later, offering flexibility and continuity in their work.
System administrators benefit significantly from ThinLinc’s robust capabilities in managing remote Linux systems. It simplifies remote server administration by offering a graphical interface that streamlines tasks such as routine maintenance, system configuration, and health monitoring, all without the need for physical presence. This remote access capability enables real-time problem-solving and efficient system monitoring, greatly enhancing administrative efficiency and reducing downtime. Additionally, ThinLinc’s secure connections and support for advanced features like GPU acceleration ensure that both administrators and users can perform their tasks with high performance and security, making ThinLinc an invaluable tool for effectively managing and utilizing remote Linux environments.
Integration with HPC Systems
Many graphical-based applications are not specifically designed for High-Performance Computing (HPC) environments due to their resource-heavy interfaces or reliance on single-threaded processing. However, running these applications in HPC settings can significantly enhance performance, particularly for scientific visualization, simulation, or large-scale data analysis. spinTwo further supports these use cases by providing robust security and seamless integration with HPC systems, ensuring that sensitive data remains protected while maximizing computational efficiency.
By utilizing ThinLinc, users can access powerful HPC resources from virtually any location, interacting with graphical applications as if they were running locally. This not only improves accessibility but also allows users to leverage HPC power without needing specialized hardware on their personal devices, enhancing both performance and convenience.
Example Use Cases
- Drug Discovery and Life Sciences: ThinLinc enhances access to tools such as MATLAB, used for data analysis and algorithm development, and the Schrödinger Suite, which supports molecular modeling and simulations. Researchers using Schrödinger’s Maestro for drug design can leverage ThinLinc to perform high-resolution molecular docking studies from any location, facilitating collaborative efforts across global research teams. Other significant tools in this field include COMSOL Multiphysics for simulating physical processes, GROMACS for molecular dynamics simulations, and Chimera for visualizing molecular structures.
- Flight Sciences: Detailed simulations and modeling are critical. Tools such as ANSYS Fluent for computational fluid dynamics and Simulink for system simulation are integral to this field. Aerospace engineers working on simulations of aerodynamics in ANSYS Fluent can access their computational models remotely through ThinLinc, allowing them to interact with complex simulations and visualize results in real-time, even when working off-site. This capability supports the development of advanced flight systems and components under various conditions.
- Gas and Oil Sector: ThinLinc optimizes the use of applications such as Petrel for reservoir modeling, Eclipse for reservoir simulation, and Hysys for process simulation. Engineers using Petrel for seismic interpretation and reservoir modeling can leverage ThinLinc to interact with large-scale geological datasets and simulation results remotely, enhancing decision-making processes and resource management. This capability is crucial for optimizing exploration and production activities.
- Academic Settings: ThinLinc offers substantial benefits for researchers and students who need access to advanced computational tools and resources. Tools such as MATLAB for numerical computing, R for statistical analysis, and specialized software for research projects can be accessed remotely through ThinLinc. For example, students and researchers using MATLAB for data analysis and visualization in coursework or research projects can access their computational environments remotely, facilitating collaboration and enhancing productivity.
HPC Integration
Graphical-based applications often struggle to fit within High-Performance Computing (HPC) environments because they rely heavily on resource-intensive interfaces and single-threaded processes. Integrating ThinLinc with HPC systems involves addressing technical challenges such as compatibility with HPC resource management, network latency, and security protocols. Effective integration requires careful configuration to ensure seamless performance and secure remote access, allowing users to interact with graphical applications as if they were local while leveraging the full power of HPC resources.
Here are simplified technical steps and command-line instructions for setting up this integration:
Install ThinLinc Server on the HPC Cluster
Ensure that ThinLinc Server is installed on the HPC cluster’s head node. Download the ThinLinc server package from Cendio’s website.
wget https://www.cendio.com/thinlinc/download/thinlinc-server-latest-x86_64.rpm sudo rpm -ivh thinlinc-server-latest-x86_64.rpm
Configure ThinLinc to Connect to HPC Nodes
Configure ThinLinc to communicate with the Slurm-managed HPC nodes. Edit the ThinLinc configuration file, typically located at /opt/thinlinc/etc/thinlinc.conf. Add the HPC nodes to the node_list section:
[cluster] node_list = headnode, gpu-node1, gpu-node2
Configure Slurm for GPU Access
Ensure that Slurm is configured to manage GPU resources. Edit the Slurm configuration file, typically located at /etc/slurm/slurm.conf, and include GPU node definitions.
NodeName=gpu-node1 Gres=gpu:4 State=UNKNOWN NodeName=gpu-node2 Gres=gpu:4 State=UNKNOWN
Install and Configure ThinLinc Client on User Workstations
Users need to install the ThinLinc Client on their workstations to access the HPC cluster remotely. Download the ThinLinc client package from Cendio’s website.
wget https://www.cendio.com/thinlinc/download/thinlinc-client-latest-x86_64.rpm sudo rpm -ivh thinlinc-client-latest-x86_64.rpm
Submit Jobs to Slurm from ThinLinc
Users can submit Slurm jobs to GPU nodes from within their ThinLinc session if the host in use is actually a Slurm submit host. This integration is made possible by configuring the ThinLinc environment to work seamlessly with the Slurm workload manager. When ThinLinc is set up on a Slurm submit host, users can leverage the cluster’s GPU resources directly through their remote sessions, making it easier to manage and execute resource-intensive tasks without needing to physically access the HPC infrastructure.
Open a terminal and use the sbatch command to submit jobs to the HPC cluster:
sbatch --gres=gpu:1 my_gpu_job_script.sh
Access GPU-based Compute Nodes
In this setup, ThinLinc acts as a bridge, providing a graphical interface that simplifies the submission and monitoring of Slurm jobs. Users can interact with the HPC cluster through ThinLinc as if they were working locally, while Slurm handles the job scheduling and resource allocation. This integration allows for efficient use of HPC resources and makes it possible to run complex computations on GPU nodes, all while maintaining the ease of use and accessibility that ThinLinc offers.
Accessing GPU-based Compute Nodes via SSH Tunneling
For users needing to interact with visualization tools on GPU-based compute nodes, SSH tunneling can be a secure and effective method to access these resources. Here’s how to set up SSH tunneling for accessing graphical applications:
Set Up SSH Tunnel
Create an SSH tunnel from your local workstation to the GPU-based node using the ssh command with port forwarding (this is done through a “proxyJump” server). For instance, to forward port 5901 (commonly used by VNC) from the remote node:
ssh -N -f -L port:localhost:port gpuNode -J loginNode
Launch Visualization Tool
After establishing the SSH tunnel, connect to the GPU node via ThinLinc and launch the visualization tool. For example, if you are using VNC to access graphical applications:
vncviewer localhost:1
Alternatively, launch the application directly if it’s capable of forwarding X11:
vmd -dispdev text -e my_script.tcl
Interact with Visualization
Once connected to the HPC cluster through ThinLinc, users can utilize the graphical interface to interact with visualization tools directly from their remote session. ThinLinc securely transmits graphical data between the GPU-based node and the user’s local workstation, allowing for real-time interaction with visualization applications. This setup ensures that even resource-intensive graphical tools can be used efficiently and securely, leveraging the computational power of the HPC cluster while maintaining the convenience and accessibility of ThinLinc as a typical Linux host graphical interface.
Accessing Visualization Tools via Jupyter Hub and Jupyter Notebooks
For data scientists taking advantage of HPC environments, ThinLinc offers an efficient solution for running Jupyter notebooks remotely on compute nodes. By initiating a ThinLinc session, users can activate a virtual environment with the necessary Jupyter modules installed and launch a Jupyter notebook server. The integrated browser within the ThinLinc session then facilitates seamless access to the Jupyter notebook interface, eliminating the need for port-forwarding or other complex configurations.
Collaborating with Cendio to Enhance ThinLinc
spinTwo’s collaboration with Cendio has significantly enhanced ThinLinc’s capabilities, especially in terms of performance and security. ThinLinc’s performance excels in graphical rendering speed, latency, and data compression when compared to other remote desktop solutions, demonstrating its superior capabilities.
In terms of security, ThinLinc has robust features that ensure data protection and compliance with stringent regulatory requirements. It offers encrypted connections and access controls that protect sensitive information, which are crucial for safeguarding data during transmission. Additionally, ThinLinc supports single sign-on (SSO), which simplifies user authentication while maintaining high security.
ThinLinc is designed to meet various regulatory standards, including those set by the Federal Information Security Management Act (FISMA) and the Health Insurance Portability and Accountability Act (HIPAA). FISMA requires federal agencies and their contractors to secure information systems, and ThinLinc’s security features align with these requirements, ensuring comprehensive protection for federal data. HIPAA compliance is essential for healthcare organizations, and ThinLinc’s encrypted communications and access controls help meet the stringent security and privacy standards required for handling sensitive health information. Similarly, in the private sector, ThinLinc aligns with industry regulations to protect proprietary data and ensure confidentiality in transactions and communications.
These advanced security features not only enhance ThinLinc’s performance but also ensure that it adheres to critical US federal government regulations, reinforcing its suitability for high-security environments. By meeting these regulatory requirements, ThinLinc enhances security and compliance, enabling organizations in regulated industries to operate securely and efficiently while leveraging advanced remote access capabilities.
Get Started with ThinLinc and spinTwo
Discover how ThinLinc can transform your remote desktop experience and enhance your HPC environment. To see ThinLinc in action or to learn more about how spinTwo’s unique blend of scientific and technical expertise can benefit your organization, contact us today. Our team is ready to provide a personalized demo and answer any questions you may have.
Reach out to spinTwo at contact@spin-two.com for more information and to schedule your demonstration. Let us help you optimize your computing environment and achieve your scientific and technical goals with confidence.
____________________________________________________________
Learn more about the partnership between Cendio and spinTwo to expand ThinLinc’s reach in HPC and secure data management across the Americas here.