   |  |
Computing Technical Design Report
| | |
Computing Technical Design Report
The technical coordination database area gathers together all the databases used for the production and installation of the ATLAS detector hardware and infrastructure. This data is used both by ATLAS technical coordination (TC) and sub-detectors during the integration and installation phases of the experiment, and will in many cases be used to initialize other databases, e.g. the online configuration and conditions databases. Some of the data will be required on a long-term continuing basis, e.g. to track the movement of equipment in and out of the experimental cavern throughout the life of ATLAS, in particular to satisfy the radiation safety requirements of the legal authorities (INB regulations).
Many separate databases have been used during the production phase of the ATLAS experiment, to keep track of detector parts, assemblies, test and quality assurance data. Much of this data is spread throughout the collaboration, using a variety of tools ranging from MS Access, MySQL and Oracle databases, to text files, reports and unstructured information. An effort is under way to migrate as much of this data as possible to the CERN Oracle physics database services, to ensure its long-term availability and accessibility. Sub-detector groups are being encouraged to create their own database schema on the CERN physics database server, with data structures appropriate to each individual data and application, rather than trying to organize all the various types of data in the same way. So far, data has been stored from a number of ATLAS sub-detectors, including mechanical characterization and electronic test data for the LAr calorimeter, and validation and certification data from several types of muon chambers.
To assist sub-detectors and TC in accessing this varied data, a generic database search interface, known as `Glance' has been developed. This system reads the database schema, and automatically produces search interfaces based on the table structure, which can be customized using XML files. This allows tailored search interfaces to be quickly set up for any set of ATLAS production data, without the need for the development of dedicated interface software. This system has been used both for production data, and the ATLAS integration and cabling databases discussed below.
Some sub-detectors have set up sophisticated database systems of their own, e.g. the SCT production database hosted on Oracle servers at the University of Geneva, used by both SCT and pixel detectors for all their module production and characterization data. In such cases, where the long-term data integrity and support is already ensured by outside institutes, no migration of production data to CERN is foreseen.
Many thousands of separate detector and infrastructure parts are being installed in the ATLAS cavern and counting rooms, all of which need to be tracked and accounted for, especially in view of the INB safety regulations. For this purpose, ATLAS is making heavy use of the MTF (Manufacturing and Test Folder) system based on web interfaces, MP5 and the CERN Oracle database servers. This system has been developed at CERN, and is also used for tracking the installation of the LHC machine. Within MTF, the ATLAS detector is represented as a tree structure, with branches for each sub-detector and the individual leaves being pieces of the detector which are installed individually - e.g. a detector module or electronics crate. Each such piece has an `equipment passport' giving its unique identifier, equipment code, physical properties (dimensions and weight), current and final locations, and links to other information such as specification documents or drawings. MTF also allows other information to be associated with each object, such as links to documents in the CERN engineering-data management system (EDMS), and in particular information on the current status of the object and tests it has undergone. All ATLAS components that are installed underground will be entered into MTF, and will be identified by barcode stickers giving the unique equipment identifier.
An important part of the installation database effort is concerned with managing the electronics racks and cables around the detector. This data is stored in the ATLAS integration database, which is maintained on the CERN Oracle physics database servers, and contains details on the location and contents of racks and the crates inside them. This database is accessed using the Rack Wizard, a Java-based graphical tool developed by CMS and in use by several LHC experiments [4-1]. The Rack Wizard is being used for defining the rack contents, planning for infrastructure needs (e.g. power and cooling requirements) and tracking installation. The system is currently being extended to cover racks and patch panels on the detector itself, and to provide graphical views of their location and contents.
A cabling database has recently been implemented in the context of the integration database framework. This database stores information on the purpose, physical characteristics, routing and status of each cable, together with the associated connectors and cable trays. It is currently being used to plan and track the cable installation, and an interface to the Rack Wizard, allowing the connection of cables to electronics boards and crates, is under development. Bulk data import and export is performed by means of Excel spreadsheets, which are prepared by sub-detector cabling experts, and search/update interfaces are under development, including access via cable-label barcode readers, which will be used to track cable installation and testing in the ATLAS cavern and counting rooms.
The cable database will also be used to track the detailed cable connectivity information from the sub-detectors, through intermediate patch panels, to the electronics racks. Cabling updates will be entered into the database using a barcode reader, and this data will then be exported to the online configuration/conditions databases, to allow channel mapping information to be automatically updated in the online and offline software. This facility will be particularly important during the commissioning phase, when many recabling operations can be expected.
A large amount of survey data is being accumulated during ATLAS integration and installation, ranging from measurements of the cavern floor displacements every few months through to the surveying of the internal geometry of assembled detector components and their final positioning in the cavern. This activity will continue during ATLAS operations, with e.g. the endcap calorimeter and muon wheel positions being resurveyed each time the detector is opened and closed. The data from such survey operations is typically processed by the CERN survey group, and made available to ATLAS in the form of reports and Excel spreadsheets. However, the data has potentially wide applicability for ATLAS reconstruction, e.g. for knowing the initial position of the muon chambers as a starting point for online and offline alignment. A survey database is therefore being designed, to centralize and make available data from survey group activities in a standardized way. This data will be stored on the Oracle physics database servers in the same way as other production and installation data, and will be made available to sub-detectors, e.g. through the Glance search interface or dedicated transfers to the conditions database.
| | |
Copyright © CERN 2005