Python for the .NET developer Transcripts
Chapter: Deploying Python web apps
Lecture: Hosting options and trade-offs
0:00 Let's spend a minute and think of the different ways
0:01 in which we could host our web app
0:04 and even amongst the different places we might choose
0:08 like Azure, or EC2, or something like AWS
0:11 something like that. There's actually a bunch of different
0:12 options in there.
0:14 One option would be to go with Azure
0:16 and we could pick several varieties in there
0:19 we could go with a software as a service
0:21 with just their web apps offering
0:23 or we could get some VMs and run things there.
0:26 Over here in EC2, or AWS, we could fire some EC2 instances
0:30 and do all sorts of cool things there
0:32 but we're going to focus specifically
0:35 on deploying straight to Linux.
0:38 So I guess that would be kind of the VM story
0:40 over at Azure, but there's better options
0:43 that than, trust me, just for the infrastructure
0:45 side of things. And I'll make a case for that.
0:48 We're going to use DigitalOcean, and we're going to
0:51 deploy to Ubuntu. Why are we going to do that?
0:53 Well put DigitalOcean to the side there for a minute
0:56 why are we going to deploy directly to Linux?
0:58 Because if you were going to deploy
1:00 to a platform as a service like Azure web apps
1:03 or Heroku, or something along those lines
1:06 you need to know a little bit about the underlying
1:09 infrastructure and how to make that work.
1:11 And you might have to configure things and so on.
1:13 To understand all of it, we're going to do it directly
1:16 on Linux, and it turns out to be actually
1:18 not very hard at all, the most affordable
1:20 and the most flexible option.
1:21 So the way I see it, if I give you the most
1:23 powerful outcome, the most flexible
1:26 and also by the way, the most affordable
1:29 by a long shot way to deploy your app
1:31 and you just decide, you know, I'd rather
1:32 put it on Azure and just push the button
1:35 to deploy there, or I'd rather just drop
1:36 it into Heroku and do a gitpush Heroku for my deploy
1:41 it's totally fine, and you'll understand
1:42 what a few of the config settings you'll have
1:44 to deal with will do, but you'll also understand
1:46 probably what's going on underneath a lot better.