Run SQL file

Run SQL file executes a script without opening it in an editor. The file streams through the statement splitter in the job runner, so large dumps run in flat memory. DELIMITER, dollar quoting and comments are handled as in the query editor, and the file’s own BEGIN and COMMIT run as written.

Run a file

1. Right-click a connection or a database in the sidebar and choose Run SQL file….
2. Choose Choose file… and pick a .sql file, or a gzip-compressed one. The dialog shows its size and the first statements.
3. Under When a statement fails, pick Stop or Log it and continue.
4. Choose Run.

Progress and failed statements show in the Jobs panel. Each failed statement is listed with its statement number, line and error message.

Run a file on a schedule

Instead of Run, choose Schedule… to run the file on a schedule. Each run executes the file as it is at that moment. On a production or confirm-writes profile, Querybara asks once, when you schedule it (Schedule it), and the runs do not ask again.

Write rules

On a production or confirm-writes profile, Querybara asks once before the run, since any statement in the file may write. On a read-only profile the run stops at the first statement that writes.

From the command line

querybara run-file runs a file the same way. By default it stops at the first failed statement; --continue keeps going, and --error-log writes the failed statements and their errors to a file. It exits 0 when every statement ran, 1 when some failed with --continue, 2 when it stopped at a failure, and 130 when interrupted.

querybara run-file dev migrate.sql
querybara run-file "postgres://[email protected]/shop" dump.sql.gz --continue --error-log errors.log

Result rows are discarded; use querybara query -f to see them.

Related

• Jobs
• Import
• Schedules
• Backup and restore