Sun Cluster Environment

Enrique Vargas brings 20 years of large systems experience to Sun Microsystems, where he currently specializes in high-end, highly available UNIX solutions. Enrique joined the Sun Enterprise Engineering group in 1997 from Amdahl Corporation, where he focused on UTS (mainframe UNIX) and high-end, highly available Solaris systems. Enrique holds a B.S. in Electrical Engineering. Joseph Bianco, Ph.D., has been working for Sun Enterprise Services for over five years as a Senior Support Engineer and has been involved with the implementation of Sun Cluster products since their inception. Prior to working for Sun Microsystems, Dr. Bianco worked for several years as a database programmer on Wall Street. Dr. Bianco holds a B.S. in Electrical Engineering and an M.S. and Ph.D. in Information Systems. David Deeths has been working for Sun's Enterprise Engineering group for over three years. David has worked extensively with existing and future Sun Cluster products. David holds a B.S. in Electrical Engineering and a B.S. in Cognitive Science.

Preface.


1. High Availability Fundamentals.


Basic System Outage Principles. System Types-Availability. Requirements. Reliability Fundamentals. Mean Time Between Failures (MTBF). Failure Rate. Common MTBF Misconceptions. Availability Fundamentals. Reliability vs. Availability. Serviceability Fundamentals. Dynamic Reconfiguration. Alternate Pathing. Single System Availability Tuning. Configuring to Reduce System Failures. Reducing System Interruption Impact. Reducing Maintenance Reboots. Configuring Highly Available Subsystems. Datacenter Best Practices. Systems Management Principles. Hardware Platform Stability. Consolidating Servers in a Common Rack. System Component Identification. AC/DC Power. System Cooling. Network Infrastructure. Security. Systems Installation and Configuration Documentation. Change Control. Maintenance and Patch Strategy. Component Spares. Software Release Upgrade Process. Support Agreement and Associated Response Time. Backup and Restore. Sun Cluster Recovery Procedures. Campus Cluster Recovery Procedures. Summary.



2. Sun Cluster 2.2 Architecture.


What Is a Cluster? Single System Image. High Availability and Scalability. Single Point of Failure (SPOF). Sun Cluster 2.2 Software Components. Cluster Interconnect. Switched Management Agent. Cluster Membership Monitor. Cluster Configuration Database. CCD Architecture. Public Network Management. Cluster Reconfiguration State Machine. Volume Management. Data Service Layer. Data Services. Fault Probes. Parallel Databases. Clustering for High Availability. Logical Hosts. Cluster Topology. Asymmetric High Availability. Symmetric High Availability. Two-Node Clusters and Cluster Pairs. N+1 Topology (Three- and Four-Node Clusters). Ring Topology (Three- and Four-Node Clusters). Scalable Topology. SC2.2 Failure Fencing/Cluster Membership. Membership and Failure Fencing. N+1 and Ring Topology. Scalable Topology. Cluster Failure Scenarios. Architectural Limitations. Summary.



3. Sun Cluster 2.2 Components.


Server Platform Options. SPARCserver 1000E (SS1000E) and SPARCcenter 2000E (SC2000E). Ultra Enterprise 1 (Ultra-1) and Ultra Enterprise 2 (Ultra-2). Netra t Carrier-Grade Servers. Sun Ultra Enterprise Servers x20R/x50. The Sun Enterprise Server xx00. Sun Enterprise 10000 Server. Solaris Operating Environment Options. Public Network Interface Options. Combination of 10/100 Base-T Ethernet and SCSI Adapter Cards. Fast Ethernet. Quad Fast Ethernet. Gigabit Ethernet. Token Ring. Asynchronous Transfer Mode. Fiber Data Distributed Interface. Private Network Interface Options. Fast Ethernet. Quad Fast Ethernet. Scalable Coherent Interconnect (SCI). Gigabit Ethernet Interconnect. Disk Storage Options. SCSI Storage Devices. Fiber Channel Storage Devices. Volume Manager Options. Solstice Disk Suite Software. Veritas Volume Manager (VxVM) Software. Veritas CVM Software. Data Storage Options. Raw Disk Storage. File System Storage. Data Service Options. Summary.



4. Sun Cluster 2.2 Administration.


Chapter Roadmap. Monitoring Sun Cluster Status and Configuration. Monitoring Tools. Cluster Logs. Managing Time in a Clustered Environment. Cluster Membership Administration. Starting a Cluster or Cluster Node. Stopping a Cluster or Cluster Node. Managing Cluster Partitions. Setting and Administering CMM Timeouts. Managing Cluster Configuration. Cluster Configuration Databases. Managing Cluster Configuration Changes. CCD Quorum. Shared CCD. Shared CCD Setup. Cluster Configuration Database Daemon. Resolving Cluster Configuration Database Problems. Administering NAFO Groups. PNM Daemon. Setting up PNM. Changing NAFO Group Configurations Using pnmset. Overview of the pnmset Options. Logical Host and Data Service Administration. Logical Hosts and Data Services. Managing Logical Hosts. Administering Logical Hosts. Cluster Topology and Hardware Changes. Administering Hardware and Topology Changes. Changing the Node Hardware. Changing Private Network Configuration (for Ethernet). Changing Private Network Configuration (for SCI). Changing Terminal Concentrator/SSP Connectivity. Changing the Quorum Device. Changing Shared Data Disks. Changing Node Topology. Summary.



5. Highly Available Databases.


High Availability for Business. Parallel Databases. Parallel Databases Using Sun Cluster 2.2 Software. Parallel Database/Highly Available Database Comparison. Failover Latency. Configuration Issues. Highly Available Databases. Minimum Configurations for Highly Available Databases. Cluster Configuration Issues Prior to Installation. Database Software Package Placement. Configuring Highly Available Databases. Creating Logical Hosts. Logical Host Syntax. HA Status Volumes. Setting Up the Logical Host Mount Point. Database Monitoring Setup. Fault Probe Account Setup. Database Monitoring. Fine-Tuning Fault Probe Cycle Times. Database Failover. Fault Probe Debugging. Handling Client Failover. Summary.



6. Sun Cluster 2.2 Application Notes.


Hardware Connectivity. Software Connectivity. Terminal Concentrator Setup. Changing the Terminal Concentrator's IP Address. Establishing a Default Route for the Terminal Concentrator. Administrative Workstation Setup. SC2.2 Client Software Installation. SC2.2 Client Software Configuration. Starting the Cluster Console. Cluster Nodes Software Installation and Configuration. Disabling the Local MAC Address. Setting the SCSI-Initiator ID. Solaris 2.7 11/99 Software Installation. Operating System-Post Installation and Configuration. Installing Recommended Solaris Operating Environment Patches. Veritas Volume Manager Version 3.0.2 Installation. Installing Recommended VxVM 3.0.2 Patches. Veritas 3.0.2 Software Configuration. Veritas VxVM 3.0.2 Post-Configuration. Sun Cluster 2.2 Installation. Sun Cluster 2.2 Post-Installation and Configuration. Failover Function Confirmation. Summary.



7. Sun Cluster 2.2 Data Services.


Highly Available Data Services. Qualifying a Highly Available Application. Client/Server Environment and Services. Crash Tolerance. Dual-Ported Data. Physical Hostname Independence. Multi-Homed Host Support. Handling Multiple IP Addresses on a Single Network Interface. The Failover Process. Dependent Data Services. Cluster Daemons and Reconfiguration. Data Service Methods. Data Service Method Registration. Data Service Method Execution Order. Data Service Method Program Arguments. Developing HA-Agents. SC2.2 Monitoring Tools. Summary.



8. Beyond Sun Cluster 2.2.


Main Features of the Sun Cluster 3.0 Software. Continuous Availability. Clustering within the Solaris Operating Environment. Operating System Clustering Advantages of SC3.0 Software. Topology. Topologies Supported by SC3.0 Software. Network Adapter Failover. Cluster Interconnects. Local Storage (Boot Disks) for SC3.0 Software. Shared Storage. Global Devices. Resource Group Manager. Data Services and Resource Types. New Cluster API. Object-based. Single System Image for Standard Applications. Global File System (GFS). Highly Available, Scalable Data Services. Types of Scalability. Communications Framework. Cluster Management. Summary.



Appendix A:


SCSI-Initiator ID. SCSI Issues in Clusters. Changing the SCSI-initiator ID Overview. Changing the SCSI-initiator ID. Using the Open Boot PROM (OBP) Prompt Method. Maintaining Changes to NVRAM. Summary.



Appendix B:


SC2.2 Data Service Templates. Starting, Stopping, and Aborting Data Services. Fault Monitoring. Summary.