Veritas Cluster Server Syllabus

Getting acquainted with Clustering

  • What is a cluster?
  • What is a VCS cluster?
  • Detecting Failure
  • Switchover and failover

Understanding cluster components

Resources

  • Resource dependencies
  • Resource categories

Service groups

  • Types of service groups
  • The cluster service group

Agents

  • The agent framework
  • Agent Operators
  • Agent Classifications

Attributes

  • Attribute data types
  • Attribute dimensions
  • Types of attributes
  • Keywords/reserved words

Cluster control, communication and membership

  • High-Availability daemon (HAD)
  • Low latency transport (LLT)
  • Group membership service/atomic broadcast (GAB)
  • Inter-node communication
  • Intra-node communication
  • Low priority link
  • Disk heartbeats (GABDISK)
  • Jeopardy
  • Jeopardy conditions
  • Examples of jeopardy and network partitions

Configuration Concepts

The VCS configuration language

The main.cf file

  • Include Clauses
  • Cluster definition
  • System definition
  • Service group definition
  • Resource dependency clause

The tyes.cf file

Managing the VCS configuration file: the hacf utility

  • Verifying a configuration
  • Loading a configuration
  • Dumping a running configuration
  • Multiple versions of .cf Files

Putting the pieces together

  • Initial configuration
  • The main.cf for a two-node asymmetric NFS cluster
  • Configuring application service groups using application wizard

Cluster topologies

Basic failover configurations

  • Asymmetric or active/passive configuration
  • Symmetric or active/active configuration
  • N-to-1 configuration

Advanced failover configurations

  • N+1 configuration
  • N-to-1 configuration

VCS user privilege model

  • VCS user privileges
  • User privileges for CLI and Cluster shell commands

Administering VCS through command line

VCS environment variables

Starting VCS

  • Stopping VCS without –force option
  • Additional considerations for stopping VCS

Changing the configuration mode

  • Setting the configuration to read/write
  • Setting the configuration to read-only

Managing cluster users

  • Adding a user
  • Modifying a user
  • Deleting a user
  • Assigning and removing user privileges
  • Displaying a user

Querying VCS

  • Querying service groups
  • Querying resource types
  • Querying resources
  • Querying agents
  • Querying systems
  • Querying clusters
  • Querying status

Administering service groups

Administering resource types

Administering resources

Administering systems

Administering clusters

Backup up and restoring VCS configuration files

  • Creating snapshots
  • Displaying snapshot information
  • Updating the snapshots
  • Hasnap –d diff
  • Exporting the snapshots
  • Modifying the snapshot list
  • Hasnap –exclude
  • Deleting the snapshots

Controlling VCS behaviour

VCS behavior on resource faults

VCS behavior diagrams

  • Scenario: resource with critical parent faults
  • Scenario: resource with non-critical parent faults

Controlling VCS behavior at the service group level

  • Controlling failover on service group or system faults
  • Controlling clean behavior on resource faults
  • Controlling fault propagation

Controlling VCS behavior at the resource level

  • Attributes to remember
  • VCS behavior when an online resource faults
  • VCS behavior after a resource is declared faulted

Disabling resources

  • When to disable a resource
  • Limitations
  • How Disabled resources affect group states

Clearing resources in the ADMIN_WAIT state

Troubleshooting VCS startup

  • Troubleshooting VCS startup
  • Troubleshooting service groups
  • Troubleshooting resources
  • System states
  • Examples of system state transitions

Performance Considerations

  • How cluster components affect performance
  • Kernel components (GAB and LLT)
  • The VCS engine "HAD"
  • The impact of agents
  • Booting a cluster system
  • Bringing a resource online
  • Taking a resource offline
  • Bringing a service group online
  • Detecting resource failure
  • Detecting system failure
  • Detecting network link failure
  • Time taken for a service group switch
  • Time taken for a service group failover
  • CPU binding of HAD
  • Monitoring CPU usage
 

Duration

30 Hrs

Download VCS Syllabus