OpenShift vs. Heroku for Python Web Apps

Should you already knowledgeable enough with internet application, I bet you know that locating a cheap (yes, cheap must come first) and dependable (subsequently come dependability) hosting service for your PHP program is very simple. It Really Is because PHP, or to be specific LAMP stack (Linux Apache MySQL PHP) is the most well-known technology stack in developing web software. A yearly cost to get a LAMP stack is very affordable. But the same is not used if you try and construct your net application with technologies stack other than LAMP, like Ruby, Java or Python. In my experience, a free strategy is important, as it enables you to get examine your program on the web at no cost. And having a VPS will need you to handle and install all your application connected framework, library and program on your own.

If you are like me, in some point in your search of a totally free hosting for the Python program, I’m sure you will stumble upon Heroku. It is a Platform as a Service (PaaS) remedy already live long enough that host many fantastic web application. It allows you to focus on your program, instead of fretting about your server installation as well as upkeep. It come with a totally free 1 Dyno, which is a device of your application example running to serve request. Their free strategy may alter as time goes by, but currently a free plan of unverified account has 5 free applications limit. Beside free programs, heroku also come with the idea of addons, where you could add any available technology to your now running application including : PostgreSQLMongoDB, MySQL, Redis, Hadoop and a lot more. Some contain a complimentary strategy, the others do not. Heroku have its Heroku toolbelt, which will be a CLI application dealing with management of your application. It is actually interesting working with Heroku!

However, in the specific context of free strategy, Heroku has an incredibly large constraint. In my opinion, it’s a deal breaker. Those are :

  • A free strategy for PostgreSQL (or may also other database, haven’t check it however), consist of a 5MB size limitation. So what can you do with just 5MB database size?
  • A totally free application may not store user uploaded file. Thus, if you strive to build a some form of Facebook killer program (go on, dream on…) your user will not be able to add their avatar.
  • Finally, a free software has 100MB quota for space size including all library, framework and program code.

In fact, I have attempted to constructed a program in Heroku, attempting to fail that 5MB constraint on PostgreSQL. But, I just can’t continue working having a totally free plan. So, my search to get a great free plan of my Python web application proceeds. And .. I ended up in OpenShift.

In expression of Platform as a Site alternative, OpenShift come with all Heroku synonyms, but using an ample free option“Three programs, each having 1Gb room quota including your theoretical model, library, program code, database and user uploaded files. In my opinion, it is possible to really create a real world alternative with only a totally free strategy. I understand, trendy isn’t? Cool.

As I said OpenShift come with all Heroku synonyms, it signifies that in OpenShift you will located exactly the same thing such as Tools (==Dyno) and Cartridge(==add ons). In addition, it include a CLI program where you can easily manage your own application. This mean, all you can do with Heroku, you can do it also with OpenShift.. only more generous.

2 thoughts on “OpenShift vs. Heroku for Python Web Apps

  1. Great post! Thanks for the comparison. I’ve looked at Heroku before, just to play around. The paid options there seemed pretty expensive. So cool to hear about OpenShift and their generous free tier. I’m signing up for an account right now.

Leave a Reply

Your email address will not be published. Required fields are marked *