Davix

Davix
Name	Davix

Davix is an open-source client library and set of command-line tools for interacting with distributed storage protocols and web-based data access services. It provides utilities for HTTP(S), WebDAV, GridFTP, and specialized high-performance transfers, enabling integration with scientific workflows, data grids, and data management systems. Davix is commonly used alongside middleware and services in high-energy physics, astronomy, and large-scale computing infrastructures.

Overview

Davix functions as a bridge between applications and storage services such as HTTP, WebDAV, and GridFTP endpoints, supporting secure authentication schemes and resilient transfer mechanisms. It exposes a C/C++ API and command-line utilities that integrate with systems like XRootD, dCache, EOS and iRODS. The project is frequently associated with middleware stacks used by collaborations such as CERN, European Organization for Nuclear Research (CERN), European Grid Infrastructure (EGI) and experiments including ATLAS and CMS.

History and Development

Davix emerged to address the need for robust data access libraries compatible with grid-era services pioneered by projects such as LHC Computing Grid, Worldwide LHC Computing Grid (WLCG), and related national research infrastructures. Early development intersected with protocols and tools like GridFTP, XRootD, and service frameworks from CERN IT teams. Over time, contributions have come from research organizations, computing centers such as Karlsruhe Institute of Technology (KIT), and national laboratories including INFN and FNAL. The codebase evolved to support modern TLS stacks and federated identity providers such as OAuth 2.0, X.509, and integration with services operated by European Space Agency (ESA) and astronomy projects like Square Kilometre Array planning groups.

Features and Architecture

Davix exposes features for high-performance transfers, protocol negotiation, and authentication. It implements connection reuse, pipelining, and multi-stream handling to improve throughput when interacting with servers such as Apache HTTP Server or NGINX. The library supports certificate-based authentication via X.509 and proxy mechanisms used in grid middleware, token-based schemes aligned with OAuth 2.0 and OpenID Connect, and integration with federated identity providers like eduGAIN. Its modular architecture allows backends for WebDAV and GridFTP while providing a uniform API for client applications such as data management tools used by CERN OpenStack, HEPIX deployments, or science gateways built atop Globus services. Davix also integrates with monitoring and logging stacks like Prometheus and ELK Stack through hooks and metrics endpoints.

Use Cases and Applications

Davix is used in workflows spanning high-energy physics, astronomy, climate science, and bioinformatics. It underlies data transfer components in pipelines for collaborations including ALICE, LIGO, and large survey projects such as Pan-STARRS and Gaia. Research facilities and computing centers like CERN, DESY, SLAC, and national supercomputing centers use Davix to enable access to tape gateways, object stores tuned for scientific workloads, and federated storage federations such as EUDAT. Integration examples include job payloads orchestrated by HTCondor, data replication driven by Rucio, and archival retrievals managed by dCache and TAR systems.

Performance and Benchmarks

Performance evaluations of Davix focus on throughput, latency, and resilience under concurrent transfers. Benchmarks often compare Davix-based transfers to native GridFTP clients, curl, and protocol-specific daemons like XRootD clients, measuring metrics such as sustained throughput over wide-area networks and behavior under high-latency conditions typical of intercontinental links used by WLCG. Optimization efforts address TCP tuning, parallel streams, and TLS session reuse to approach performance of dedicated high-throughput tools such as bbcp and FDT. Community reports from computing centers like CERN IT and NCSA provide empirical throughput and failure-mode statistics for production deployments.

Adoption and Community

Davix adoption is concentrated in scientific computing communities tied to major research infrastructures and experiments. Contributors and users include software teams from CERN, INFN, KIT, DESY, and national research networks such as GÉANT and Internet2. The project is hosted in open-source ecosystems where collaboration overlaps with tools like Rucio, HTCondor, Globus, and iRODS. Governance and contribution follow typical open-source models with issue trackers and mailing lists; collaborations with initiatives such as Open Science Grid and training events at conferences like CHEP help sustain the developer and user base.

Security and Compliance

Davix implements secure transport through TLS and supports authentication methods including X.509 certificates and token-based frameworks like OAuth 2.0. Deployments frequently adhere to institutional policies from entities such as CERN and national data protection regulations like GDPR where applicable. Integrations with identity federations such as eduGAIN and authorization frameworks enable compliance with access control practices used by data stewardship frameworks in projects like EUDAT and EOSC. Security assessments performed by computing teams emphasize certificate management, revocation handling, and secure defaults for cipher suites compatible with guidance from organizations including IETF working groups.

Category:Open-source software