Cloud9 VS Codeanywhere - What do you think about cloud IDEs?

#1

I like the idea of cloud IDE for collaboration and version control. I’m also tired of apps working fine in the local environment but failing on the staging server. What I am extremely interested in is setting up an automatic deployment workflow to Galaxy from the cloud. We have a tiny team so the goal is to collaborate with experienced devs as fast as possible, wherever they are, whatever their environment needs are.

What do you guys think? Let’s have a chat!

P.S Codeanywhere will support docket soon - https://codeanywhere.com/features/devBox/#container

#2

Since AWS bought Cloud9, I’ve been using it and like it. I had a free C9 account before but it didn’t have premium features like ssh.

I already use a few EC2 instances for work, its much more convenient to share with colleagues and to run heavy jobs in the cloud vs my pc. Since we use S3, its much much faster to do that from within an EC2 running in the same network, and there’s no cost.

Cloud9 on AWS is free when I use it via ssh (which was a paid feature earlier). It works great.

2 Likes
#3

Thank you @dirkgently! Sounds pretty good! How is your deployment workflow process? Do you directly deploy from Cloud9? Using CLI? Are you hosting exclusively with AWS? I am interested in keeping everything in Galaxy which runs on AWS so hopefully, I can figure out an easy deployment workflow

#4

Yes we are hosted almost all on AWS now, we also have maybe 2 $5 droplets on DO for minor tasks because its cheaper. I am in the process of converting most of our server infrastructure to lambda because you get 1m calls/month free which is more than enough for a startup.

Code is hosted on Visual Studio Online github (unlimited free repos and lots of features. BitKeeper can’t compare, CodeCommit is a joke).
Storage is in S3.
We use MongoDb’s Atlas for the DB (internally its on AWS as well)
There are 2-3 Meteor servers running in EC2. I am also working on using autoscaling so we don’t pay for extra servers.
Deployment right now is done via mup (its the easiest I’ve found). When I get autoscaling working we will get rid of that as well.
I also have Kadira running on another EC2 box.

Code can be done locally or on EC2 via Cloud9. I use cloud exclusively these days since I can login, resume where I left off, and I can run/debug on the same instance. So e.g. to run from within Cloud9 terminal I can do ‘yarn start’, or to deploy ‘mup deploy’

Galaxy would cost much more, and it doesn’t include DB anyway.

1 Like