Ahmed Ejaz
|
067fa80d0f
|
Merge pull request #13458 from rioug/docker-upgrade-postgres
Docker upgrade postgres
|
2025-10-15 02:24:16 +05:00 |
|
Gareth
|
6f9dcf7e27
|
remove bundle check
|
2025-08-20 11:04:40 -04:00 |
|
Gareth
|
2d064bab64
|
Created a bundler service that runs once removing responsibilities from any other services. The bundler service always runs install which should still be pretty fast if nothing or only a few gems have changed. A healthcheck won't work unless bundler runs continuously which is impractical. Instead, a checksum is generated on bundle install and sentinels in the other services have a definite confirmation that bundle is complete. The nice thing about this approach is that web, webpack, and sidekiq (which share the same bundles dependencies) will not concurrently run bundle install solving dep install redundancies.
|
2025-08-20 10:13:20 -04:00 |
|
Gareth
|
b95d798a27
|
Fixed webpack service so that web relies on its bundles. This has re-enabled JS
|
2025-08-18 12:24:27 -04:00 |
|
Gareth
|
e1e4aeac1f
|
Added conditions to sidekiq too as it races db as well. Everything seems functional now
|
2025-08-18 11:52:52 -04:00 |
|
Gareth
|
c7ae47053e
|
Added health check to avoid docker container racing
|
2025-08-18 11:38:57 -04:00 |
|
Gaetan Craig-Riou
|
d3a2c09f66
|
Upgrade to postgres 16
Postgres 16 is shipped with Ubuntu 24 which is the version we will be
moving to in production
|
2025-08-18 10:45:54 +10:00 |
|
Gaetan Craig-Riou
|
d596e692d8
|
Upgrade postgres to version 14
This is the version currently availablein Ubuntu 22
|
2025-08-18 10:45:24 +10:00 |
|
Gaetan Craig-Riou
|
2df95dcbab
|
No need to expose the postgres port
There is no need to expose the port to the host, all container will be
running in the same network, so they will have access to database port.
|
2025-08-18 10:45:18 +10:00 |
|
Gareth
|
74368f939b
|
By creating the db in the container on composition, the first migration file rejects the schema of the empty database. All I had to do was remove db creation from docker so that db:prepare will default to creating open_food_network_dev from schema.rb rather than perceiving a mismatch from the precreated docker db
|
2025-08-15 14:23:13 -04:00 |
|
Gareth
|
3750898c44
|
Looks like db:prepare never fully ran and that's why it did not work. The issue seems to be using foreman with web, webpack, and sidekiq in the same script. Though not dependent on each other, the build order or port assignment was causing web to exit early upon build
|
2025-08-06 15:21:30 -04:00 |
|
Gareth
|
d34f8900d7
|
divided foreman queued services into 3 containers. Web was exiting seemingly at random and seems to be a conflict between web, web-pack, and sidekiq being run through foreman. The division into 3 dev containers has been very consistent in building the project correctly
|
2025-08-06 15:05:45 -04:00 |
|
Gareth
|
6a912b7d8c
|
Added cmake dep to dockerfile and added script for db:schema:load that runs only if the schema is different than the latest migration
|
2025-07-30 20:12:53 -04:00 |
|
François Turbelin
|
0fdf789249
|
Clean remaining config for Docker system specs
|
2025-02-04 22:59:31 +01:00 |
|
François Turbelin
|
a6666dd061
|
Use only new Dockerfile
|
2025-01-26 22:15:46 +01:00 |
|
François Turbelin
|
a1c95e1837
|
Use new Dockerfile in Docker Compose
|
2025-01-26 21:44:07 +01:00 |
|
François Turbelin
|
39822e1761
|
Launch new Docker procfile from Docker Compose
|
2025-01-19 22:17:31 +01:00 |
|
François Turbelin
|
2acde5423d
|
Remove extra containers
|
2025-01-19 21:00:28 +01:00 |
|
Neal Chambers
|
078e191d26
|
Sync Docker Container to the Host
|
2024-11-28 08:39:51 +09:00 |
|
Neal Chambers
|
02e2214caa
|
Revert "Remove bundle exec from docker commands"
This reverts commit 2d193a689406cf826e241314acd661fd87a6ae37.
|
2024-11-28 08:39:51 +09:00 |
|
Neal Chambers
|
efe0a2a701
|
Add Redis Test URL for Spree Preferences
|
2024-11-28 08:39:51 +09:00 |
|
Neal Chambers
|
5c5213e872
|
Use Alpine Image for Docker
|
2024-11-28 08:39:51 +09:00 |
|