run

Synopsis

starlake run [options]

Description

Execute the project's tasks in dependency order, in parallel, inside this JVM. The graph is built from transform lineage, so a load table is executed only when some transform reads it: tables no transform references are never ingested by this command. When a transform and a declared load table share a domain.table name, the two collapse to a single node and the transform is what runs, never the load.

Selector syntax, shared by --select and --exclude:

• domain.table - exactly that task
• domain.* - every task in the domain
• tag:VALUE - every task carrying the tag
• +expr - the matched tasks and all their transitive upstreams
• expr+ - the matched tasks and all their transitive downstreams
• +expr+ - both directions

Matching is case-insensitive. A run executes the selected set and nothing else: unselected upstreams are not run implicitly, so a task whose input is missing fails normally. A selection that matches no task exits with code 3.

Parameters

Parameter	Cardinality	Description
--parallelism <value>	Optional	Max concurrently executing tasks. Defaults to the maxParTask setting (SL_MAX_PAR_TASK). Raising it runs tasks concurrently through code paths that are not concurrency-safe in all engines, so treat it as an explicit opt-in.
--fail-fast <value>	Optional	Abort the whole run on first failure instead of only the failed branch
--select expr1,expr2...	Optional, Unbounded	Selector expressions, unioned. Repeatable, and each value may itself be a comma-separated list. Omit to select the whole project.
--exclude expr1,expr2...	Optional, Unbounded	Selector expressions subtracted after --select. Exclusion wins.
--dry-run <value>	Optional	Resolve and print the execution plan, execute nothing
--options k1=v1,k2=v2...	Optional	Variables passed to the templating path
--reportFormat <value>	Optional	Report format: console, json, html