This is the parent issue to track progress on the schemachange workload project. The goals and motivation of the project are discussed in detail in this doc.

Error Classification

• Add knob to control the frequency of errors (workload/schemachange: improve errors; initial changes #55521)
• Classify and handle errors (workload/schemachange: classify and handle errors in all operations #56119)

Complex Schema Generation

• Support user defined schemas (workload/schemachange support user defined schemas in workload #54889)
• Create views/tables that depend on other views/tables (workload/schemachange: improve errors; initial changes #55521)
• Unique constraints (workload/schemachange: add fk constraints and type improvements  #58065)
• Single FK constraints (workload/schemachange: add fk constraints #57472)
• Composite FK constraints
• Check constraints
• Serial column types that are backed by sequences
• Use DEFAULTs in insertRow op. This will require small updates in unique/fk constraint checking where we will have to look up column defaults and sub them in.

Observability

• Improve logging and add backups workload/schemachange: add logging and backups  #57471

Automation

• Improve github issues created by roachtest failures roachtest: automate github issues on schemachange random load failures #58245

Other Improvements

• Handle aborted transactions / update nested function calls to use nested transactions (workload/schemachange: handle aborted transactions #56120)
• Use better knobs workload/schemachange: knob improvements #57300
• Have a way to debug errors in the workload workload/schemachange: add mechanism to detect op gen failures #57949

Jira issue: CRDB-3577