ShaoLin InfiniCluster 2.0 User Guide
PrevNext

I. Basic Concept and Terminology

Table of Contents
1. Overview of Cluster Technology
1.1. What is a Cluster?
1.1.1. Failure Detection
1.1.1.1. Application Failure Detection
1.1.1.2. Node Failure Detection
1.1.2. Switchover, Failover and Failback
1.1.2.1. Switchover
1.1.2.2. Failover
1.1.2.3. Failback
1.2. How to make Application "Clustered"?
1.2.1. Start, Stop and monitor procedures
1.2.1.1. Defined Start Procedure
1.2.1.2. Defined Stop Procedure
1.2.1.3. Defined Monitor Procedure
1.2.2. Known state
1.2.3. Store required data on shared disks
1.2.4. Adherence to system
2. General hardware configuration
2.1. General hardware configuration
2.2. Configuring Storage
2.3. Configuring SCSI storage
2.4. Configuring single initiator SCSI
2.5. Configuring multi initiator SCSI
2.6. Configure SCSI Host Bus Adapters (HBAs)
2.7. Configuring SCSI IDs
2.8. Configuring Fibre Channel
2.9. Configuring power supply for storage system
2.10. Internal Storage
2.10.1. Advantages
2.10.2. Disadvantages
3. Introduction to ShaoLin InfiniCluster
3.1. Terminology
3.2. Cluster Features
3.3. Understanding ShaoLin InfiniCluster Components
3.3.1. Resources
3.3.1.1. Resource Group
3.3.2. Service Monitoring Agents (SMA)
3.3.2.1. Agent Architecture
3.3.2.1.1. Agent Operations
3.3.2.2. Agent Types
3.3.3. Cluster Infrastructure and Control
3.3.3.1. The ShaoLin General Parallel Cluster Infrastructure (SGPCI)
3.3.3.1.1. Membership
3.3.3.1.2. Heartbeat
3.3.3.2. Load Balanced Communication
3.3.3.3. Reliable Data Communication
3.3.3.4. Kernel Watch Dog and Kernel Kill
3.3.4. Cluster Manager
3.3.4.1. Cluster Configuration Database (CCDB)
3.3.4.2. Cluster Administration Daemon (CAD)
3.3.4.3. Group Communication Service (GCS)
3.3.4.4. System Resource Management Service
3.3.4.5. SIGKILL and Cluster Manager
3.4. Cluster Administration and User Interface
3.4.1. Command Line Interface
3.4.2. Graphical User Interface (GUI)
PrevHomeNext
ShaoLin InfiniCluster 2.0 User Guide Overview of Cluster Technology