Website's previous version is available at www60.jinr.ru

Seminar of Alushta-2025 scientific conference participants

Meshcheryakov Laboratory of Information Technologies

Joint Laboratory Seminar

Date and Time: Wednesday, 3 June 2026, at 11:00

Venue: room 310, Meshcheryakov Laboratory of Information Technologies

  1. Seminar topic: «Prototyping the SPD Online Filter as a Private Cloud Infrastructure: Towards a Dynamical System Formulation»

    Speaker: Nikita Greben

    Аннотация:

    The SPD Online Filter is an essential intermediate system between the continuous data stream from the free-running DAQ and the distributed offline data processing system for the SPD experiment at NICA. Acting as a sophisticated software trigger, the Online Filter must handle high data rates, necessitating a highly scalable and resilient online filtering middleware. We present an Infrastructure as Code (IaC) pipeline for the automated provisioning, configuration management, and orchestration of a private cloud infrastructure. Beyond practical deployment, a reproducible infrastructure provides a key foundation for studying the system’s complex dynamics. Finally, we describe how the private cloud can serve as a flexible testbed, enabling multifactorial debugging of both middleware and application software within the SPD Online Filter.


  2. Seminar topic: «Data Flow and Data Control in the SPD Online Filter Middleware»

    Speaker: Polina Korshunova

    Аннотация:

    The Spin Physics Detector (SPD) is being constructed as a part of the NICA mega-scientific project at the Joint Institute for Nuclear Research. It is expected that the data flow from the detector may reach up to 200 PB per year due to the absence of a hardware trigger. To reduce the volume of data for long-term storage and subsequent analysis, the primary processing will be performed by a specialized computing system – SPD Online Filter.

    This report describes the design and integration of two components developed for the SPD Online Filter: the DAQ Emulator Service and the Data Management System (DMS). The DAQ Emulator is responsible for data flow modeling. The recorded data primarily ends up in the DMS, which is responsible for managing the lifecycle of experimental data.

  3. Seminar topic: «SPD Computing Infrastructure and Data Production»

    Speaker: Alexey Konak

    Аннотация:

    Preparations for the SPD experiment at the NICA collider are underway. While the detector is being built, large-scale Monte Carlo simulations and data reconstructions are already being performed. The process of obtaining data ready for final analysis is a complex multi-stage processing, at each step of which intermediate data is generated. A special infrastructure has been created to organize the complete production chain, as well as the processing and storage of these large amounts of data. This report describes the current status of the SPD Computing System, and data generation process. It explains the data production process, how production tasks are organized, and how the system coordinates the production process.


  4. Seminar topic: «From a Static Pipeline to a Self-Managing One: A Workflow Management System for the SPD Online Filter»

    Speaker: Leonid Romanychev

    Аннотация:

    The SPD experiment at the NICA collider produces enormous data streams that cannot be stored as-is. Before anything reaches long-term storage, the SPD Online Filter reduces this flow through a chain of automatic processing stages. Coordinating that chain — deciding which step runs next, where its results are placed, and what can be safely discarded — is the job of the Workflow Management System (WfMS). This talk presents the WfMS developed for the SPD Online Filter. The system does not blindly follow a predefined pipeline: it reacts to its own environment — taking into account how full the data storage is, how long each task has been waiting, and the role the task plays in the processing chain — and uses this information to decide what to run next. It also cleans up after itself — removing unused intermediate data automatically and handling failures gracefully — which makes the pipeline more stable and reduces routine manual work for operators. The result is a system that manages itself, better prepared for the sustained data rates expected from the SPD Online Filter.


  5. Seminar topic: «SPD Production control panel»

    Speaker: Nikita Monakov

    Аннотация:

    The SPD (Spin Physics Detector) is currently being built as part of the NICA complex at JINR. Simultaneously with detector construction, the experiment’s software ecosystem is being developed to support large-scale simulation of physical processes.

    According to the Technical Design Report, the collaboration expects high luminosity, which will result in substantial data volumes. To support such a workload at scale, we have developed a specialized system that automates routine operations and consolidates all actions required to configure both individual tasks and task chains in a single interface.

    The system is already operational and has helped physics working groups launch dozens of tasks to simulate millions of events. This talk presents the design and implementation of the production control panel system.


  6. Seminar topic: «Development of additive technologies and 3D modeling at MLIT JINR»

    Speaker: Andrey Evlanov

    Аннотация:

    The report discusses the main stages of the creation and development of the MLIT JINR additive technologies and 3D modeling laboratory. Special attention is paid to the integration of high-tech equipment into the institute’s infrastructure to solve engineering problems. The report describes the experience of using various 3D modeling methods to create modified power supply systems that ensure the operation of MICC JINR IT equipment. The article describes the main stages of the additive technologies and 3D modeling laboratory development, 3D modeling methods, and the prospects for scaling up within the framework of MLIT JINR development.

By using the JINR web-site, you accept the cookies that we use. Learn more about how we use cookies.