1 / 1

Next-Generation Navigational Infrastructure and the ATLAS Event Store

Algorithm (e.g. Track Finder 1). Algorithm (e.g. Track Finder 2). Algorithm (e.g. Jet Finder 1). Converter. Next-Generation Navigational Infrastructure and the ATLAS Event Store. POOL/ROOT File. Tools. Link Id. Cont. Token. Store Gate. Persis- tency. <#>. Cont ,….

isra
Télécharger la présentation

Next-Generation Navigational Infrastructure and the ATLAS Event Store

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript


  1. Algorithm (e.g. Track Finder 1) Algorithm (e.g. Track Finder 2) Algorithm (e.g. Jet Finder 1) Converter Next-Generation Navigational Infrastructure and the ATLAS Event Store POOL/ROOT File Tools Link Id Cont. Token Store Gate Persis- tency <#> Cont,… Services Abstract: • The ATLAS event store employs a persistence framework with extensive navigational capabilities. These include real-time back navigation to upstream processing stages, externalizable data object references, navigation from any data object to any other both within a single file and across files, and more. The 2013-2014 shutdown of the Large Hadron Collider provides an opportunity to enhance this infrastructure in several ways that both extend these capabilities and allow the collaboration to better exploit emerging computing platforms. Enhancements include redesign with efficient file merging in mind, content-based indices in optimized reference types, and support for forward references. The latter provide the potential to construct valid references to data before those data are written, a capability that is useful in a variety of multithreading, multiprocessing, distributed processing, and deferred processing scenarios. This paper describes the architecture and design of the next generation of ATLAS navigational infrastructure. ATLAS Software & Event Data Model: The ATLAS experiment at CERN uses the Athena software framework to run simulation, reconstruction, and analysis. StoreGateallows software modules to transparently use a data object that was created by an upstream module or read from disk. The Athena software framework: Uses a transient store to exchange data and interface to I/O and persistence infrastructure. POOL / ROOT Persistency Framework: Athena software uses ROOT I/Ovia the ATLAS POOL persistency framework, which provides high-performance and highly scalable object serialization to self-describing, schema-evolvable, random-access files. Input File Persistent State (P) Transient State (T) Baskets (B) Compressed baskets (b) The Athena persistency framework: Reading POOL/ROOT data via the Athena software framework read decompress stream t/p conv. • References / Token: • ATLAS I/O and persistence infrastructure supports extensive navigational capabilities: • Real-time back navigation to upstreamprocessingstages. • Externalizabledata object references. • Navigation from any data object to any other both within a single file and across files. Token ##Links OID1 OID2 POOL Token references and Link table: Tokens store the storage technology, database identifier, persistent class identifier, container name, and two offsets ContName, OID2 Enhancements to the Navigational Infrastructure: References to event entry points will be designed to use unique, immutable {run number, event number} pairs. ROOT support for content-based indexing: ROOT supports an in-file indexing capability that provides the foundation for efficient implementation of content-based references to event data in ATLAS. Because ATLAS event numbers will be 64-bit integers, a number of enhancements are required to ROOT indexing and to its underlying formula evaluation infrastructure. Immutable references for merged files: For event data stored in ROOT, merging files means appending the Tree’s entries. But Token references require the offset identifiers to be updated. The use of immutable attributes in the Token will allow re-indexing to be handled by the underlying storage technology. Support for forward references: In practice, to construct an event reference to a standard ATLAS data file one needs only to identify the file and to specify which event, as all other attributes have predefined default values. File identifiers can be pre-generated, allowing ATLAS to define forward references to events in downstream data products. P van Gemmeren1, D Malon1 and M Nowak2on behalf of the ATLAS Collaboration 1Argonne National Laboratory, Argonne, Illinois 60439, USA 2Brookhaven National Laboratory, Upton, New York 11973, USA E-mail: gemmeren@anl.gov

More Related