Importing & Restoring

crdb-dump load applies schema and loads data via CockroachDB COPY.

Schema + data

crdb-dump load --db=mydb \
  --schema=crdb_dump_output/mydb/mydb_schema.sql \
  --data-dir=crdb_dump_output/mydb \
  --validate-csv

• --schema applies a .sql DDL file (statements are split safely, respecting string literals and function bodies).
• --data-dir loads every *.manifest.json found in the directory.
• --validate-csv checks each chunk's header against the live table columns before loading.

Parallel loading

crdb-dump load --db=mydb --data-dir=crdb_dump_output/mydb --parallel-load

Dry run

crdb-dump load --db=mydb --data-dir=crdb_dump_output/mydb --dry-run

Resumable loads

Loads record progress so re-runs skip already-loaded chunks (idempotent):

# single shared log file
crdb-dump load --db=mydb --data-dir=... --resume-log=resume.json

# one log per table
crdb-dump load --db=mydb --data-dir=... --resume-log-dir=resume/

# abort on the first failed chunk
crdb-dump load --db=mydb --data-dir=... --resume-log-dir=resume/ --resume-strict

Selecting tables on load

crdb-dump load --db=mydb --data-dir=... --include-tables=mydb.public.users
crdb-dump load --db=mydb --data-dir=... --exclude-tables=mydb.public.audit_log