meteor-up: 502 Bad Gateway When Trying to Use LetsEncrypt/Autogenerate
I get a 502 Bad Gateway when I include the autogenerate section of ssl below. If I comment it out and redeploy, everything works great. When I load the page, it does load with the cert over https, but it’s 502ing. Running docker logs bxlp-nginx-letsencrypt --tail 500 looks like it has the cert and says it does not need to renew it.
Here are my mup logs:
[138.197.112.225]
[138.197.112.225]> fibers@1.0.15 install /bundle/bundle/programs/server/node_modules/fibers
[138.197.112.225]> node build.js || nodejs build.js
[138.197.112.225]
[138.197.112.225]`linux-x64-46` exists; testing
[138.197.112.225]Binary is fine; exiting
[138.197.112.225]
[138.197.112.225]> meteor-dev-bundle@0.0.0 install /bundle/bundle/programs/server
[138.197.112.225]> node npm-rebuild.js
[138.197.112.225]
[138.197.112.225]
[138.197.112.225]> bcrypt@1.0.2 install /bundle/bundle/programs/server/npm/node_modules/bcrypt
[138.197.112.225]> node-pre-gyp install --fallback-to-build
[138.197.112.225]
[138.197.112.225]node-pre-gyp ERR! Tried to download: https://github.com/kelektiv/node.bcrypt.js/releases/download/v1.0.2/bcrypt_lib-v1.0.2-node-v46-linux-x64.tar.gz
[138.197.112.225]node-pre-gyp ERR! Pre-built binaries not found for bcrypt@1.0.2 and node@4.7.2 (node-v46 ABI) (falling back to source compile with node-gyp)
[138.197.112.225]make: Entering directory '/bundle/bundle/programs/server/npm/node_modules/bcrypt/build'
[138.197.112.225] CXX(target) Release/obj.target/bcrypt_lib/src/blowfish.o
[138.197.112.225] CXX(target) Release/obj.target/bcrypt_lib/src/bcrypt.o
[138.197.112.225] CXX(target) Release/obj.target/bcrypt_lib/src/bcrypt_node.o
[138.197.112.225]/opt/meteord/run_app.sh: line 21: 8 Killed npm install --unsafe-perm
Here’s my mup.js:
module.exports = {
servers: {
one: {
host: 'IPADDRESS',
username: 'root',
pem: 'KEYHERE'
// password:
// or leave blank for authenticate from ssh-agent
}
},
meteor: {
name: 'NAMEHERE',
path: 'PATHHERE',
servers: {
one: {}
},
buildOptions: {
serverOnly: true,
},
env: {
ROOT_URL: 'https://DOMAIN.COM',
MONGO_URL: 'mongodb://localhost/meteor'
},
dockerImage: 'abernix/meteord:base',
ssl: {
autogenerate: {
email: 'EMAIL',
domains: 'SUB.DOMAIN.COM'
}
},
deployCheckWaitTime: 60
},
mongo: {
oplog: true,
port: 27017,
servers: {
one: {},
},
},
};
About this issue
- Original URL
- State: closed
- Created 7 years ago
- Reactions: 1
- Comments: 34 (14 by maintainers)
@maxmatthews I needed this fix today. Thank you. After upgrading to a 1gb digitalOcean droplet, everything worked great. It really cut down on the random fails I’d get once in a while with the smaller droplets.
@zodern are the minimum requirements specified anywhere in the readme? It seems if autogenerate is being used, a 1gb seems to be the minimum requirement for RAM. Using ubuntu 14.04, etc. I’d be happy to write a “recommended minimum requirements” section for the readme and send you a PR? I could at least start the section for my experiences on both AWS and DigitalOcean.
@ivanthemeh Make your own issue and post the link here and we can jump on that thread. Will make it easier for more people to follow both our issues as I think they are separate. Happy to help you once you do that.
@maxmatthews sorry, I was answering @ivanthemeh. Did you try to add indexes to your Mongo database. A friend of mine was having serious CPU overload (it went up to 120% with just a few users) and it decreased down to ~10% when he added indexes on Mongo. It was caused by some composite publications
Maybe you can try that, see if that works
Does this not work on ubuntu 16? I am having the same issues as you except Im using an instance with 1gb of ram…