To deploy venome we use the Oregon State Center for Quantitative Life Sciences (CQLS) servers. They have a linux virtual machine that we can ssh into and expose our application to the internet.
1. Make an account through CQLS
First you need to have an account through CQLS. So first sign up https://shell.cqls.oregonstate.edu/access/ to get access. I'd reccomend you make the CQLS username just your ONID one.
Once you make an account, ask Nate or Michael to email the admin (Ken) about adding you to our venome linux VM. Otherwise, you won't have access.
2. ssh
You'll need to ssh into a few servers to get in. your_assigned_port is an integer and should be given to you upon registration of your CQLS account through email. your_cqls_username is simply your CQLS username you entered when registering.
First
ssh -p your_assigned_port your_cqls_username@shell.cqls.oregonstate.eduand enter your CQLS password
Then once you are in
ssh hpc.cqls.oregonstate.eduand enter your OSU ONID password, not your CQLS password this time.
Then finally
ssh venome-pvtand enter your OSU ONID password again.
✅ Success you are in!
When you do
lsYou should see the venome project already there. If you don't, that's because the user home/bb/bertuccd/ originally cloned the repo and ran the docker. Please contact the admin so that these files can be shared in a group with you too. As of now there was no better solution other than one person cloning and running the code. You might need to come up with a better shared solution if nothing comes up.
If you need to reclone the repo, make sure to use the https link and the not ssh link for the github repo link. No clue why but the ssh link hangs forever.
For some reason, if you are on a proxy server, docker can't connect to the internet at all. To fix this, you need to point docker to what the $http_proxy enviroment variables says with https://docs.docker.com/config/daemon/systemd/. Follow those instructions. You might also need to update the docker/config.json with the same info. In my case there was a config.json in ~/.docker/config.json and in /etc/docker/config.json. No clue why there are multiple, but just make sure they all agree and restart the daemon and docker.
You'll need to restart docker and the daemon if you change the config with sudo systemctl daemon-reload and sudo systemctl restart docker.
Go to the etc/httpd/conf/ directory and find the httpd.conf configuration file.
Add a reverse proxy so that the server directs user traffic to our running production build
ProxyPass /backend/ http://localhost:8000/
ProxyPassReverse /backend/ http://localhost:8000/
ProxyPass / http://localhost:5173/
ProxyPassReverse / http://localhost:5173/You might need to sudo vim to actually be able to edit these files. Once you do this restart the apache server with sudo service httpd restart.
For deployment, we use the production build scripts. Please see build.md for how to run the docker container in production mode.
You first need to change the frontend/buildConstants.ts variable named BACKEND_URL to the URL your frontend will be requesting to. In this case change it to https://venome.cqls.oregonstate.edu/backend since that what the frontend will need to call to when deployed.
Then you can build the entire container in production
./run.sh start -p
./run.sh reload_from_backup -p backups/v0.0.2 # or whatever backup you wantOn the server you can simply make a backup with any name you want like new_backup as
./run.sh backup -p backups/new_backupFirst you need to make a backup which I described in the last section, then logout of the venome-pvt shell connection. From your local machine make sure you are either on the Oregon State Internet or through a VPN. Then
sftp your_cqls_username@hpc.cqls.oregonstate.eduThen download the specific backup (in this case downloading v0.0.2, but could pick any one)
get -R venome/backups/v0.0.2I would save this to a central box or google drive so that you can save backups over time.