Plan for the use of SAM-Grid for Reconstruction

Oct 11, 2004

Participants

FNAL: Gabriele, Anoop, Bimal | Daniel, Mike, Amnon (?) | Iain, Peter Love (?)

Remote Sites: Tibor, Joerg, …

WBS

1. Development

1.1. JIM

1.1.1. Client

1.1.1.1. Definition of the JDL (Prototype Done)

1.1.1.2. Consistency Checks (Prototype Done)

1.1.2. Exec site

1.1.2.1. Coordinating data retrieval from SAM (Done)

1.1.2.2. Interaction with RUNJOB (Prototype 80% Done)

1.1.2.3. Output sandbox (Prototype Done)

1.1.2.4. Merging (By Oct 29)

1.1.3. Monitoring via XMLDB (Prototype Done)

1.1.3.1. Adding job status information in JIM and RUNJOB

1.1.3.2. Displaying the info on the web

1.1.4. Submission site

1.1.4.1. Fault tolerance policies

1.1.4.2. Structured job management

1.1.5. Brokering site

1.1.5.1. Resource selection algorithms

1.2. RUNJOB

1.2.1. Environment preparation for Reco

1.2.1.1. For old (test) DZero Code Version (p14.05.01) (Done)

1.2.1.2. For p17

1.2.1.3. For merging (By Oct 29)

1.2.2. Store data for Reco

1.2.2.1. Metadata (Prototype 80% Done)

1.2.2.2. Durable Location (By Oct 15)

1.2.2.3. Merging (By Oct 29)

1.3. Test

1.3.1. Development at samgfarm

1.3.1.1. Unit tests (Prototype Done)

1.3.1.2. Grid tests (By Oct 13)

1.3.2. Integration

1.3.2.1. Deploy at FNAL-FARM (By Oct 13)

1.3.2.2. Stress test (By Oct 22)

1.3.2.3. Generate development items overt feedback (By Oct 29)

1.3.3. Pre-production

1.3.3.1. Deploy and test at other sites (Wisc., Man., …) (By Oct 29)

1.4. Refinement / Feedback integration

2. Deployment

2.1. Manual deployment

2.1.1. Upgrade SAM-Grid sites (By Nov 19)

2.1.2. Install new sites

2.2. Assisted deployment

2.2.1. Refine samgrid automatic installation software

2.3. OSG compatibility

2.4. Documentation

2.4.1. Educate stakeholders

2.4.2. Refine installation instructions for reco (By Oct 29)

2.4.3. Write deployment documents

3. Maintenance / Operations

3.1. Migration to v6 (in 2005)