How to setup your PostgreSQL database template
Our DB templates for PostgreSQL come with a build-in initialization script that allows you to seed you DB during the build process. This allows you to ‘bake’ your seed into your master DB container. Then every branch that you run in isolation will immediately start up with a DB container that is pre-seeded, taking advantage of the container image cache.
- Make sure you have set up your database template first.
Create a Non-Repository Template for PostgreSQL.
- Click Create Template in the Templates tab
Select Non-Repository Template to reveal a list of non-repository templates. Select PostgreSQL.
Name your template, if desired. Click Create Template
Your template should launch and turn green when the container has finsihed building.
Select Dockerfile to customize your database. Here you can modify the environment variables to change desired usernames, passwords, and database names. Defaults have been set for your convienence, but if you would like to change anything, just uncomment the line and modify as desired.
- Once running, you can setup template connections with your repository applications. Here is how.
Produce a dump of your development database to upload to Runnable
- Take a dump of your PostgreSQL database. The following command will dump the contents of your database named
Note the use of
-U postgres, you will need to authorize the pg_dump command with whichever custom credentials you have set.
$ pg_dump -U postgres -Fc postgres -f seed.dump
- Your dump file should end up looking similar to this, but much longer:
$ cat seed.dump PGDMP postgres126.96.36.199.8`0ENCODINENCODINGSET client_encoding = 'UTF8'; falsea00 STDSTRINGS STDSTRINGS(SET standard_conforming_strings = 'on'; falseb12621214postgreDATABASExCREATE DATABASE postgres WITH TEMPLATE = template0 ENCODING = 'UTF8' LC_COLLA TE = 'en_US.utf8' LC_CTYPE = 'en_US.utf8'; DROP DATABASE postgres; ...
Upload the seed file to your template configuration
- Bring up the template configuration modal by either clicking on the cog next to your database template in the
Containerspane, or clicking on Dockerfile in the
- Select the
Dockerfiletab on the top menu within the modal.
- Click on the
[+]button above the file explorer to the left of the Dockerfile editor.
- Select ‘Upload File’ to bring up the file selector and choose your dump file (i.e.
Enable the database seeding functionality
- Uncomment the
ADD seed.dump /seed/dumpline in the Dockerfile editor to the right.
If you have named your seed something other than
seed.dump, you should modify the first part of the add command. Make sure you leave the second
seed.dumpintact so the init script will work.
- Now click
Save & Buildto build your container and seed your database.
Now that your template database container has been ‘baked’ in with your seed file, subsequent databases that are spawned per branch will immediately start up with the seeded data. If you ever want to update your seed file, just click
Rebuild Without Cache on the template container to pull a newer seed and cache it into the container image.
Want to seed your database from S3? Here is how.