07 · Tables, End to End (A–Z)

This section walks the entire life of a table - from an empty bucket to a retired table - as one journey. The stages are lettered A–N; read them in order the first time, or jump to the stage you need.

The journey

• A · Before you start - the mental model, fully-qualified names, the stable UUID, prerequisites.
• B · Plan the table - namespaces, column types, partitioning, naming.
• C · Create the table - create_table, folders, a demo table, verifying.
• D · Design the schema - infer, inspect, add columns, tune settings.
• E · Govern with column domains - fixed, hierarchical, and reference domains; validate.
• F · Load data - import a CSV, insert, merge, first-load checklist.
• G · Read & explore - query, SQL, aggregate, profile.
• H · Edit data over time - append, update, delete, bulk corrections.
• I · Version, time-travel & recover - history, diff, time-travel, rollback, restore.
• J · Organize & reshape - rename, move, manage files.
• K · Maintain & tune - compaction, optimize, vacuum, cadence.
• L · Row & column security - Cedar policies that filter rows and hide columns per user.
• M · Share & audit - share links, access, the audit log.
• N · Retire - drop a table (irreversibly) and clean up.

First time here?

The fastest way to learn the tools is the demo table - seed demo_products, then run through reading, domains, and versioning against it. You'll need a storage connection first.

Two things that recur

• Batch your writes - one merge_rows of many rows is one version, not many.
• Validate after every bulk change - validate_constraints + validate_domain_values.