Berkeley Open Infrastructure for Network Computing (BOINC)

• Introduction

Developing distributed applications with BOINC

• Projects
• Compute model: files and file references
• Compute model: applications, platforms and versions
• Compute model: workunits
• Compute model: results
• Compute model: batches
• Compute model: work distribution
• Compute model: handling very long computations
• Compute model: remote file access
• Accounting and result validation
• Back end examples
• The API
• The graphics API
• Application development

Road map of the BOINC software distribution

• Road map

Operating a BOINC project

• Installing a BOINC server
• The master URL
• Project startup checklist
• The BOINC database
• The BOINC scheduling server
• Operational tools: security
• Operational tools: applications and versions
• Operational tools: work and results
• Security

The BOINC client

• Core Client: file structure
• Core Client: FSM structure
• Core Client: data structures
• Core client: logic
• Core client: debugging

Participating in BOINC projects

• Participant accounts
• Preferences
• Participant startup

Implementation and debugging notes

• Test applications and scripts
• The BOINC client/server protocol
• Scheduling server: policy
• Scheduling server: implementation
• Scheduling server: debugging
• Host measurements
• Scheduler RPC timing and retry policies
• Persistent file transfers