Added simple dockerfile and compose #59
Conversation
* Support .env file * Fix default sub endpoint
* Setup migrations, misc tidy/fixes * Simplify migration provider
readme tweaks
* update README with instructions for the default URL * Update README.md --------- Co-authored-by: Daniel Holmgren <[email protected]>
* describeFeedGenerator route + multiple feeds * tweak readme
* describeFeedGenerator route + multiple feeds * tweak readme * improve env
* describeFeedGenerator route + multiple feeds * tweak readme * improve env * publish script * create -> put * readme * handle blob encoding * add check that feeds are available
So readers won't have to google Bluesky PDS to try to figure out what it means
(cherry picked from commit 1b2f04a)
* improve docs & helpers around publisher dids * add docs to env
Cosmetic: remove trailing space
|
Naive question: What does better control entail? |
I'm running the feed generator on my own server with lots of other stuff. If I instead of running yarn/node directly on the server but in a container I have more control of the security. Now if there is a security issue somewhere in the feed generator code, it will have impact on the container, not my server. |
|
Ah I see , thanks for this. I too have been running a feed generator on a VPS and have also experienced certain conflicts/issues from running yarn/node directly on the server. I think this PR would be helpful for running a feed generator and I think it'll make the process work the same way the PDS service does. |
|
Yes please! In particular this would make feeds more plug-n-play with services like Digitl Ocean's app platform. |
haileyok
force-pushed
the
main
branch
2 times, most recently
from
d36deef to
fadd462
Compare
| FROM node:20-alpine | ||
|
|
||
| WORKDIR /app | ||
| COPY package.json . |
There was a problem hiding this comment.
| COPY package.json . | |
| COPY package.json yarn.lock . |
| WORKDIR /app | ||
| COPY package.json . | ||
|
|
||
| RUN yarn install |
There was a problem hiding this comment.
| RUN yarn install | |
| RUN --mount=type=cache,target=/usr/local/share/.cache/yarn/v6 yarn install |
target=/root/.npm for npm
https://docs.docker.com/reference/dockerfile/#example-cache-go-packages
|
|
||
| # Bare neccesary | ||
| COPY src/ ./src | ||
| COPY .env . |
There was a problem hiding this comment.
| COPY .env . |
Some values in the .env file might be secrets, so you should mount them at runtime instead of put them in the built and published image.
Using env_file in the compose file instead.
| image: feed | ||
| ports: | ||
| - "3000:3000" | ||
| restart: unless-stopped |
There was a problem hiding this comment.
| restart: unless-stopped | |
| restart: unless-stopped | |
| env_file: .env |
I wanted to run my feed as a docker service for better control.
Created a simple Dockerfile and a matching docker-compose.yml that seems to to the trick