EGEE-II: SA1 Activity

Introduction

The SA1 activity is the part of the EGEE-II project that is responsible for providing the EGEE grid services.  This responsibility includes middleware deployment, as well as day to day operations and user support.  A lot of responsibility for these activities within the regional federations rests with the Regional Operations Centres (ROCs).  The ROCs also provide operations support - acting as Grid Operations Centres in analogy with Network Operations Centres, and provide operational and performance monitoring, troubleshooting, etc. as well as providing general grid services such as VO-related services, database services, etc.  The SA1 operations activity is coordinated by the Operations Coordination Centre (OCC) at CERN.

Relationship between SA1 and LCG Operations

As described in the project proposal and technical annexe, the existing and running WLCG infrastructure is the initial instance of the EGEE-II production grid infrastructure.  The existing grid deployment activity of WLCG becomes the heart of the EGEE-II operational infrastructure.  From the start of the project new applications (Virtual Organisations - VOs) are encouraged to join and start using the EGEE-II/WLCG grid (initial contact is through the NA4 activity which is responsible for this coordination).  New Resource Centres (RCs) are also encouraged to join the infrastructure, in this case the first point of contact for both deploying the middleware and subsequent operational support is through the local Regional Operations Centre (ROC). 

The EGEE Grid Services

EGEE runs several services:

• A production service - this is initially based on the LCG-2 service, with new resource centres and new applications encouraged to participate.  This is the service that must be run as a stable, well-supported infrastructure, running only well-tested and reliable middleware.  This is where applications will run their production work, and should expect (eventually) 24x7 support.

• A pre-production service - run in parallel with the production service, but in a way that is as much like a production service as possible.  In general this is where the next version of the middleware is first deployed, where applications test that it is useable, and that the site operators verify that it meets their needs too.  Support for this service will in general be 8x5 but exceptions may be made.

• Additional test-beds are operated by various partners and organisations.  These will include:

  • A facility for demonstrations, tutorials, training, etc.  This will be in conjunction with the NA3 (training) activity.

  • Certification test-beds - run by the OCC/CIC at CERN, but including extensions managed by the ROCs for local certification as required.

  • Application test-beds may be provided by NA4 to allow application communities to do basic integration testing before even deploying on the pre-production service.

The certification and deployment process is illustrated in the figure: