Home | Legals | Data Protection | Sitemap | KIT

Data Services Integration Team

 

 

 

 

 

DSIT, the Data Services Integration Team, has the mission to provide tools, services and processes required for the uniform access to computing and storage resources. This includes access to high performance storage, replication, organisation and archival of data. Uniform access regardless of the scientific discipline using the service will ensure an efficient provisioning and support for all participating datacentres.

Seven tasks have been identified which structure the work in accomplishing the DSIT mission.

Federated Identity Management

  • to authenticate users using the identity provided by their respective home institution
  • to authorize members of globally dispersed communities
  • to account them for services used

 

Federated Data Access

  • to provide access to the same data using several access methods
  • to access replicated data in an efficient manner
  • to move data to the preferred institution

Meta Data Catalogues & Repositories

  • to facilitate efficient searches
  • to support users organising their data
  • to efficiently support administrative decisions

Archive Service

  • to safely archive important measurements or results
  • to ensure data integrity over long periods of time
  • to establish a working funding theme for data archival

Monitoring

  • to find bottlenecks in data transfers between tape and ram
  • to analyse workflows from the data perspective
  • to charge users based on the type of storage in use

Data Intensive Computing

  • to analyse, convert and preserve data
  • to optimise computing with respect to data constraints
  • to integrate data and computing aspects

 

 

 

 

The approach to provide a uniform access to several different scientific disciplines is beneficial to both scientists making use of the infrastructure and also IT service providers. The latter benefit from a decreased amount of specifically tailored systems, because more users can be served with similar tools. Much larger is the scientific benefit. Scientists from different disciplines can easily exchange and correlate their data, which may lead to new discoveries. Long term data archives will allow to apply new methods to old data and spark novel discoveries without building expensive measurement devices.