So I've thoroughly followed the Docs for both Strapi and Heroku (with MongoDb Atlas already connected to Strapi), and I might have missed something important. When I deploy the app, the site loads for a bit then crashes. The heroku logs --tail command shows this:
2020-05-08T14:41:59.625984+00:00 app[web.1]: (node:23) Warning: Accessing non-existent property 'findOne' of module exports inside circular dependency
2020-05-08T14:41:59.626067+00:00 app[web.1]: (node:23) Warning: Accessing non-existent property 'remove' of module exports inside circular dependency
2020-05-08T14:41:59.626134+00:00 app[web.1]: (node:23) Warning: Accessing non-existent property 'updateOne' of module exports inside circular dependency
2020-05-08T14:42:04.000000+00:00 app[api]: Build succeeded
2020-05-08T14:42:29.627137+00:00 app[web.1]: [2020-05-08T14:42:29.626Z] debug ⛔️ Server wasn't able to start properly.
2020-05-08T14:42:29.627879+00:00 app[web.1]: [2020-05-08T14:42:29.627Z] error Error connecting to the Mongo database. Server selection timed out after 30000 ms
2020-05-08T14:42:29.640717+00:00 app[web.1]: npm ERR! code ELIFECYCLE
2020-05-08T14:42:29.640989+00:00 app[web.1]: npm ERR! errno 1
2020-05-08T14:42:29.641812+00:00 app[web.1]: npm ERR! back-end-4@0.1.0 start: `strapi start`
2020-05-08T14:42:29.641895+00:00 app[web.1]: npm ERR! Exit status 1
2020-05-08T14:42:29.641997+00:00 app[web.1]: npm ERR!
2020-05-08T14:42:29.642078+00:00 app[web.1]: npm ERR! Failed at the back-end-4@0.1.0 start script.
2020-05-08T14:42:29.642155+00:00 app[web.1]: npm ERR! This is probably not a problem with npm. There is likely additional logging output above.
2020-05-08T14:42:29.649687+00:00 app[web.1]:
2020-05-08T14:42:29.649822+00:00 app[web.1]: npm ERR! A complete log of this run can be found in:
2020-05-08T14:42:29.649879+00:00 app[web.1]: npm ERR! /app/.npm/_logs/2020-05-08T14_42_29_643Z-debug.log
2020-05-08T14:42:29.758124+00:00 heroku[web.1]: State changed from starting to crashed
2020-05-08T14:42:30.872518+00:00 heroku[router]: at=error code=H10 desc="App crashed" method=GET path="/" host=hidden-headland-71392.herokuapp.com request_id=2eb6e39a-875b-4b2c-91cd-3b2bdd914561 fwd="81.128.178.133" dyno= connect= service= status=503 bytes= protocol=https
2020-05-08T14:42:31.888706+00:00 heroku[router]: at=error code=H10 desc="App crashed" method=GET path="/favicon.ico" host=hidden-headland-71392.herokuapp.com request_id=d96ec78a-83ea-48ca-b576-0ad62d0a15da fwd="81.128.178.133" dyno= connect= service= status=503 bytes= protocol=https
My server.json files in all 3 folder, development , production and staging are as follows:
{
"host": "${process.env.HOST || '0.0.0.0'}",
"port": "${process.env.PORT || 1337}",
"production": true,
"proxy": {
"enabled": false
},
"cron": {
"enabled": false
},
"admin": {
"autoOpen": false
}
}
I'm unsure about the Procfile, as it stands its a file in my base directory with no extension but with the line:
web: node strapi start
And for reference my package.json looks like this:
{
"name": "back-end-4",
"private": true,
"version": "0.1.0",
"description": "A Strapi application",
"scripts": {
"develop": "strapi develop",
"start": "strapi start",
"build": "strapi build",
"strapi": "strapi"
},
And my database.json file looks like this:
{
"defaultConnection": "default",
"connections": {
"default": {
"connector": "mongoose",
"settings": {
"uri": "${process.env.DATABASE_URI}",
"database": "${process.env.DATABASE_NAME}"
},
"options": {
"ssl": true
}
}
}
}
FWIW - I had similar issues. I thought I had whitelisted all IPs on Mongodb but hadn't. This step isn't included in the video referenced above and - based on the output logged to your console - I'd guess you're having the same issue.
This setting can be found in Mongodb under Security > Network Access > Add IP Address > If you use 0.0.0.0/0 it will whitelist anything Heroku throws at it.