Skip to content

Latest commit

 

History

History
170 lines (139 loc) · 5.66 KB

deploying-to-production.md

File metadata and controls

170 lines (139 loc) · 5.66 KB

Deploying to Production

Lancer uses a simple server model. All you need with Lancer is:

  • A domain
  • A server with node.js and a file system

This means you don't need anything of the following, which are required by modern app / serverless hosting:

  • A service for uploading files (S3, Google Cloud Storage, etc.)
  • A service for hosting your database (Heroku postgres, Amazon DynamoDB, serverless-capable db hosting, etc.)
  • A service for caching (memcache, redis, etc.)

Avoiding all these saves you time and money.

Depending on your setup, you can even host multiple Lancer websites on the same box, saving even more time and money.

Render.com

  • After signing up, click New Web Service
  • Connect and select your git repo
  • Fill out name
  • Open the Advanced section
  • Click Add Environment Variables
    • Set key SESSION_SECRET with a value from runing lancer secret
    • Set key LANCER_DATA_DIR with value /var/lancer-data
    • Set key LANCER_BUILD_DIR with value build
  • Click Add Disk
    • Name it something like "Lancer Data"
    • Enter /var/lancer-data
    • Pick your starting GB size depending on the needs of your project (you can increase it later)
  • Click Create Web Service

After your app deploys, push your local data to production:

  • lancer files:push my-app.onrender.com

Next, create a user for you and your client:

Custom domain

  • Click Settings
  • Scroll to bottom and click Add Custom Domain
  • Follow instructions
  • Wait for Render to generate SSL certificates for you
    • This could take a while, even after the UI says it's complete

Digital Ocean (CapRover)

CapRover Setup

If you've already installed CapRover, you can skip to app setup

App Setup

  1. Visit your CapRover dashboard
  2. Navigate to Apps in the sidebar
  3. Check the Has Persistent Data checkbox in the form
  4. Fill out a name for your app
  5. Click Create New App
  6. Scroll down and click your new app
  7. Navigate to the App Configs tab
  8. Add some environment variables: a. Key SESSION_SECRET with a value from running lancer secret locally a. Key LANCER_DATA_DIR with value /usr/data/lancer
  9. Scroll down and click Add Persistent Directory
  10. Enter /usr/data/lancer for Path in App and something like my-app-lancer-data for Label
  11. Click Save & Update

Next we will set up deployment. In your local terminal, create a file named captain-definition at the root directory of your project (if you aren't using yarn, replace yarn with npm and remove yarn.lock):

{
 "schemaVersion": 2,
 "dockerfileLines": [
    "FROM node:14-buster",
    "RUN mkdir -p /usr/src/app && mkdir -p /usr/data/lancer",
    "WORKDIR /usr/src/app",
    "COPY package.json yarn.lock /usr/src/app/",
    "RUN yarn && yarn cache clean --force",
    "COPY . /usr/src/app/",
    "RUN yarn build",
    "ENV PORT 80",
    "EXPOSE 80",
    "CMD [ \"yarn\", \"start\" ]"
  ]
}
Alpine version, may run into dependency problems ```json { "schemaVersion": 2, "dockerfileLines": [ "FROM node:14-alpine", "RUN apk update && apk upgrade && apk add --no-cache git", "RUN mkdir -p /usr/src/app && mkdir -p /usr/data/lancer", "WORKDIR /usr/src/app", "COPY package.json yarn.lock /usr/src/app/", "RUN apk add --no-cache --virtual .gyp python make g++ && yarn && yarn cache clean --force && apk del .gyp", "COPY . /usr/src/app/", "RUN yarn build", "ENV NODE_ENV production", "ENV PORT 80", "EXPOSE 80", "CMD [ \"yarn\", \"start\" ]" ] } ```
Multistage alpine version, but very slow until caprover supports multistage caching ```json { "schemaVersion": 2, "dockerfileLines": [ "FROM node:14-alpine as builder", "RUN apk add --no-cache git python make g++", "RUN mkdir -p /usr/src/app", "WORKDIR /usr/src/app", "COPY package.json yarn.lock /usr/src/app/", "RUN yarn", "COPY . /usr/src/app/", "ENV NODE_ENV=production", "RUN yarn build", 
"FROM node:14-alpine as app",
"RUN mkdir -p /usr/app && mkdir -p /usr/data/lancer/cache/build",
"WORKDIR /usr/app",
"COPY --from=builder /usr/src/app/node_modules /usr/app/node_modules",
"COPY --from=builder /usr/src/app/data/cache/build /usr/data/lancer/cache/build",
"COPY . /usr/app/",
"ENV NODE_ENV=production PORT=80",
"EXPOSE 80",
"CMD [ \"yarn\", \"start\" ]"

]

}
</details>

**Add this new `captain-definiton` file to git**, commit, and then run `caprover deploy`.

Next, go back to your CapRover app dashboard:

- Navigate to the `HTTP Settings` tab
- It should tell you your app's subdomain. Click it to check if the deploy worked.
- In the `HTTP Settings` tab, click `Enable HTTPS`
  - This will generate a certificate for you
  - Once done, visit the https version and see if it worked
  - If it worked, scroll down and click the `Force HTTPS...` checkbox
  - Click `Save & Update`
- Optional: Connect a separate domain to your project in the same `HTTP Settings` tab

Once your site is running, you can push up your local data:

- `lancer files:push example.com`

**Recommended:** Add a `deploy` script to your `package.json` that runs `caprover deploy --default`