Software

Saba Sehrish
Fermilab

Software Co-leader

Kevin Lannon
University of Notre Dame

Software Co-leader

Overview

The Software group develops and supports software that enables CMS data processing. This software is roughly divided into two categories: Computing Infrastructure and Services software facilitates data processing by moving data and scheduling computational tasks to distribute data processing across the globally distributed CMS computational resources. Framework and Applications software provides the physics algorithms, detector simulation, and the supporting framework that processes CMS data.

2.1 Computing Infrastructure and Services

• 2.1.1 Data Management Metadata refers to the services that track metadata for CMS data handling. The two serices are the CMS Data Bookeeping Service (DBS) and the Data Aggregation System (DAS). Both pieces of software provide facilities for locating data and tracking metadata.
• 2.1.2 Monitoring refers to the collection of software, web services, and infrastructure to facilitate monitoring, both in real time and historical, of CMS data processing.
• 2.1.3 Rucio software manages data transfers across the globally distributed CMS computational infrastructure. Rucio is community-supported software used across a number of HEP experiments. U.S. CMS is leading the transition of CMS to Rucio to achieve more powerful and flexible data management capabilities.
• 2.1.4 GlideinWMS is the software resposible for provisioning computing resources for CMS. It creates a global pool of computational resources and handles matching of computational tasks with the appropriate resources on which to execute them.
• 2.1.5 Workflow Management software converts abstract workflows into sets of computational tasks that can be executed with the CMS global pool. This software is responsible for coordinating and controlling the computational work of data processing as well as tracking its progress.
• 2.1.6 XRootD software provides the capability to stream data from storage at one computational site to a processing task at a different site. This capability increases the flexibility of the CMS distributed computational infrastructure.
• 2.1.7 Web Services Deployment is the infrastructure facilitating the deployment of web services in support of data processing. The underlying technology involves Docker containers orchestrated via Kubernetes.

2.2 Framework and Applications

• 2.2.1 Framework CMSSW framework provides the ability to write reconstruction and analysis applications in a multi-threaded-capable environment with an improved efficiency.
• 2.2.2 FWLite known as Framework-light enables users to write analysis code in ROOT sessions with CMS data formats and libraries available.
• 2.2.3 Profiling refers to the effort related to profiling automation of the workflows in collaboration with the CMS Computing Modernization working group, and ensuring that CMSSW code is compliant with modern C++ standards, and can compile and run on advanced hardware with new architectures.
• 2.2.4 ROOT CMS uses ROOT to make data objects persistent. ROOT provides all the functionalities needed to deal with big data processing, statistical analysis, visualization and storage.
• 2.2.5 Simulation Geometry corresponds to the detector geometry maintenance and upgrades to provide detector geometry information to the simulation and reconstruction applications of CMS. The most critical ongoing work is migration to DD4HEP for Run3.