mirror of
https://github.com/jbranchaud/til
synced 2026-01-03 15:18:01 +00:00
Add Seed Production Data Into Another Branch as a Planetscale TIL
This commit is contained in:
28
planetscale/seed-production-data-into-another-branch.md
Normal file
28
planetscale/seed-production-data-into-another-branch.md
Normal file
@@ -0,0 +1,28 @@
|
||||
# Seed Production Data Into Another Branch
|
||||
|
||||
When you [create a Planetscale
|
||||
branch](https://planetscale.com/docs/reference/branch) off `main`, it will only
|
||||
copy over the schema. No data will be copied over to that new branch.
|
||||
|
||||
You can copy data over from the initial branch (`main`) in two steps from the
|
||||
CLI. First, create a dump of the branch. Then restore the dump into your new
|
||||
branch.
|
||||
|
||||
```bash
|
||||
$ pscale database dump database-name main --output ./dump
|
||||
```
|
||||
|
||||
That creates SQL files locally in the `dump` folder with both schema and data
|
||||
statements.
|
||||
|
||||
That folder of SQL files can then be restored into one of your branches.
|
||||
|
||||
```bash
|
||||
$ pscale database restore-dump database-name branch-name --dir ./dump --overwrite-tables
|
||||
```
|
||||
|
||||
The `--overwrite-tables` flag is needed because your branch's existing schema
|
||||
will conflict with the `create` schema statements in the SQL files.
|
||||
|
||||
You can `pscale shell` into that branch and run a `select ...` statement to
|
||||
check out the data.
|
||||
Reference in New Issue
Block a user