Research Team: Upgrading the DMS to enable Distributed Computing
Greetings NuNetopians,
We recently released a blog explaining our newly formed Research Team within NuNet. This specialized team is dedicated to advancing the frontiers of decentralized computing, focusing on both the core platform and innovative use-cases. NuNet Research is organized into distinct areas of focus, each contributing to the evolution of our platform. The purpose of this team is to research integrations and enhancements to the NuNet platform with the goal of improving features and functionalities.
In this article, we will delve into the details of the DMS upgrade, outlining the key components and goals that will transform how we manage and execute distributed computing tasks. This upgrade is not just a technical enhancement; it’s a stride towards realizing our vision of a more efficient, flexible, and powerful distributed computing environment.
The term DMS refers to ‘Device Management Service’, which is essentially the NuNet platform itself. It is the lightweight binary which establishes secure peer-to-peer connection between all computers on-boarded onto the network.
Key Research Components of the DMS Upgrade
There are several key components designed to integrate traditional machine learning libraries and improve job distribution management. The following points outline the necessary enhancements:
- Integration of IP over LibP2P: The DMS should support IP over LibP2P, enabling peer-to-peer connectivity essential for leveraging distributed machine learning frameworks like Horovod. This feature will facilitate peers within the distributed network to connect and communicate more efficiently, thereby enhancing the distributed computing capabilities.
- Enhanced System and Networking Configuration: Additional configurations for system and network are necessary to empower the DMS containers. These configurations will enable the containers to distribute computing jobs seamlessly after they are received from the service providers. It is crucial to ensure that the DMS containers can manage job distribution without bottlenecks.
- Peer Targeting Capability: The upgrade should include a feature that allows service providers to target specific peers through the service provider dashboard. This will provide service providers with the flexibility to select optimal peers for job execution based on various parameters like computational power, location, or availability.
- Job Propagation to Multiple Peers: When a service provider requests the execution of multiple jobs for distributed computing, the DMS should have the functionality to propagate these requests to multiple peers. This feature is vital for parallel processing and efficient job execution across the distributed network.
Goals of Upgrade
The enhancements aim to:
- Foster efficient peer-to-peer connectivity for distributed computing.
- Implement advanced configurations for better job distribution and management.
- Provide service providers with targeted peer selection capabilities.
- Enable the DMS to handle multiple job requests simultaneously for effective parallel processing.
The impact of these upgrades will be profound, significantly elevating the performance and reliability of our distributed computing tasks. Service providers will experience unprecedented control and flexibility in job management, leading to optimized resource utilization and enhanced efficiency.
Milestones
Our journey towards this upgrade involves several milestones, from integrating IP over LibP2P to developing peer targeting functionality. These milestones are supported by resources like technical documentation, coding structure access, UI/UX design expertise, and testing tools.
As we embark on this collaborative effort to upgrade the DMS, we invite our community to join us in witnessing and contributing (through GitLab/Discord) to this exciting phase of advancement in our distributed computing capabilities.
Stay tuned as we progress and share updates on each milestone, and join us in shaping the future of decentralized computing with NuNet.
NuNet Is Hiring!
NuNet currently has a number of open positions for various roles within the team. If you have the skills and desire to join us in our journey, you can find more information and contact us through our career page.
About NuNet
NuNet lets anyone share and monetize their computing resources, turning cloud computing power from a centralized service into an open protocol powered by blockchain. Find out more via:
