Benefits of Workshop Learning: a Cassandra Workshop Example

Cassandra Workshop

Day One:

• Introduction and discussion of workshop success criteria
• Customer overview of a few planned application uses of Cassandra
• Team discussion picking two to three applications to drill down on

Day Two:

• Summarize Day One
• Drill down on each of the selected applications to discuss:
  • Business requirements
  • What data will be stored
  • How data will be written and retrieved from Cassandra
  • Table formats and Cassandra features appropriate to each use case

Day Three:

• Summarize Day Two
• Review Day Two conclusions and re-visit each
• Operational and best practices discussion Part One:
  • Overview of Cassandra architecture and operational implications
  • Overview of Cassandra features, including dos and don’ts

Day Four:

• Summarize Day Three
• Operational and best practices discussion Part Two:
  • Key spaces
  • Overview of table configuration options and operational impacts:
    • Compaction strategies
    • Compression strategies
    • Caching options
  • Server configuration:
    • Operating System
    • Java (cassandra-env)
    • Cassandra YAML:
      • • Seeds
        • Snitches
        • Backup
        • Thread pool configuration
    • Security considerations:
      • • SSL
        • Cassandra user security
        • Using Cassandra user security with JMX
    • Monitoring and metrics
    • Irregular maintenance processes
      • • Adding new nodes
        • Removing nodes
        • Adding a new DC
        • Upgrading Cassandra versions
        • Dealing with corrupt SSTables

Day Five:

• Summarize the week
• Cover parking lot topics
• Review and evaluate success criteria

Post Workshop:

• Workshop report including:
  • Summary of the discussions throughout the workshop
  • Final recommendations on data models and query design
  • Operational best practices:
    • OS tunable settings
    • Cassandra configuration settings
    • Backup and restore recommendations
    • Repair scheduling
    • Monitoring and metrics

Ongoing pre-launch support including:

• Assistance with cluster construction
• Setting up regular repair
• Setting up backup/restore processes
• Cluster health check
• Ad hoc requests