passenger / website under heavy load - ruby-on-rails

running passenger v 5.0.11 on ubuntu, behind nginx, with ruby 2.1.3p242
The app is doing a concurrent call (generate a pdf file via whthlktopdf, opening the app one some url and saving it to a file) when a specific model is updated
not saying this is a good pattern, thing is the call get stuck and the passenger process reaches its passenger_max_request_queue_size
hence I'm wondering if there would be a way to kill stuck processes automatically when this kind of thing occurs ?

There is a builtin feature for that: https://www.phusionpassenger.com/library/config/nginx/reference/#passenger_max_request_time

Related

Dockerized Rails 5 RC1 application not picking up updates to controllers and models in development

I have quite a bit of experience developing Rails 4 apps on Mac OS X + Docker Machine + Docker Compose, but something has changed with how Rails 5 is caching files in the development environment (currently testing with RC1).
After starting the application with docker-compose up, the application runs normally in development mode.
But if I make a change to a controller or model, the only way I can get that reflected in the application is to stop the server and start it back up.
So now my workflow looks something like this when I need to make a change to a controller or model:
Make change to controller/model class.
Stop server with Ctrl + C.
Start server back up with docker-compose up.
Wait on the server to start up.
Run whatever I was running in the browser.
To say the least, needing to do steps 2-4 is annoying and not what I'm accustomed to from Rails 4.
I went a step further and uninstalled Spring using the Removal instructions, but I still get the same behavior.
I also searched for any settings that I could find in config/environments/development.rb related to the Rails runtime's iron grip on the model and controller classes, and I couldn't find anything. (I assume that it really wants to rely on watching the filesystem in order to selectively load changes, and something isn't being picked up from a change on Mac to VirtualBox.)
Any other ideas on what I can try? Or are there any new settings related to how this more aggressive caching works?
I'm Using Rails 5 rc1.
Rails 5 introduced some "improvements" to code reloading, but it doesn’t seem to work with Docker on OS X.
So in config/environments/development.rb, I replaced
config.file_watcher = ActiveSupport::EventedFileUpdateChecker
with
config.file_watcher = ActiveSupport::FileUpdateChecker
It seems that for the ActiveSupport::EventedFileUpdateChecker file watcher, the change event does not occur for docker-machine shared files

OSX Local Server Setup

I recently setup a local server on my mac. I have rails and ruby and rack installed and running.
But I'm unsure of how to direct incoming request or have a rails app take control.
How do I move past just having the index page show and have pages like http://10.0.0.2/example/test.html or whatever load???
Any resources is much appreciated, or even a couple keywords to start googling
Easiest is to install passenger and use passenger preference pane to configure it.
http://www.fngtps.com/passenger-preference-pane

Deploying Rails as a desktop application

I just start RoR development for two weeks, It's pretty amazing programming language so far. It's impressing me so much so that I'm start to thinking about change the programming language of my client's project from PHP to RoR. But what I want to know is, is it possible to make a RoR web app looks like a Native App when using it in a local machine. By the term of Native app I mean ..
The user doesn't have to run script/server or rails server every time they starts their machine
Either access it by Web Browser or something else is fine.
Can be running with Window 7 or Window XP
Other recommendation are welcome. (Including, the reason not to use RoR to develop a local application)
I personally don't think there's anything ridiculous with wanting to deploy a Rails app as a desktop application but it's certainly not the easiest thing to do. Heres a list of what you're going to need:
a compiled distributable ruby interpreter
a distributable database
a script to bootstrap the whole thing( preferably that runs on the click of an icon or when system starts)
an installer
optionally:
write a simple shell in qt or java(possibly with the ruby bindings even) that handled starting and stopping the rails app, and provided a webframe for your customer to use instead of the standard web browser.
Here's a tutorial on deploying a rails app as a desktop app, though for debian, that should start you in the right direction. But your best bet for a quick gui application in ruby is shoes.
since rails is framework for web development there is no way to avoid starting rails server to show something on local machine. you might hide the fact that server is being started by running it as service without letting the user know about it but that's only as far as you can get.
so yes, it is possible to bundle your application into a "native app" but i'm afraid it will be extremely fragile as there are too many pieces that need to fit together: ruby itself, database to use, rails and all the other gems. big chance is you will have a gem that needs to be compiled and then you're in trouble.
bottom line: it's not worth it, either host your app or pick a better solution to develop something that runs only locally.
Maybe out of topic, but if you want to make native application, perhaps make native application? I mean, what's the purpose to develop a web application and then host it locally?
At least, you could consider using free hosts, such as heroku. http://your-project-name.heroku.com will grant you a 5mb database and 1 worker. It's free, will save you a lot of troubles and will make it easier for you to deploy newer version to your clients.
Just have them put a bookmark on their desktop if they want to double click to open application.
Maybe I am wrong but you seem to miss the fact that RoR is not a programming language. Ruby is a programming language and RoR is web-development framework. Anyway if you want to develop desktop applications, you can use Ruby with a framework such as Bowline (to keep the MVC spirit).
If you want a Web application, build a Web application. If you want to build a desktop application, try Monkeybars or Appcelerator.
Quite late for my answer but anyway I had the same problem and use the gem ocra to build an exe for windows.
It is also possible to generate a setup to install the program in the register.
It is still necessary to have a server started but the user has just to click on the exe program; A console is opened and give some instructions like 'after the start of the server, open a browser and go to localhost:3000; Ctrl + C to exit when you have finished.
It could be possible to have the server launched as a service but I think my users prefer to know if the program is running or not.
ocra is not really easy to set but eventually it works fine.

Is there a way for a ruby on rails app to support web-based upgrades like Wordpress does?

I've been using Wordpress for awhile, it's installed on my own server, and one of the features I most love about it is how every time there is a new version it alerts you within the app's web-based admin and with one click I can upgrade the app. I don't have to get anywhere near a console.
Personally I wouldn't mind updating manually, but I can see how this could significantly affect the adoption of a piece of software. I'm working on creating a full-featured ruby on rails forum software and I would love to figure out how to include this feature. Any ideas if this could be done with rails?
Could a rails app modify it's own files? If it did, would the server need to be restarted?
To complicate things further, what if the app was deployed from a repo. Could the rails app check in a commit of itself after updating?
Maybe packaging the core of the app as a gem would be simpler? Then maybe the upgrade would not actually modify the rails MVC stack (the rails app would just be super-basic), instead if the forum was all contained within a gem then all it has to do is trigger a 'gem update [name]'. If this occurred, I don't think the Gemfile would even need to be updated. Would a server restart even be required to load the updated gem?
Ideas or feedback on any of this?
Rails files can be modified and even deleted on production - in my case aplication is still working unchanged as all classes are cached in memory. It means Rails instances must be restarted to take new change.
I suppose WordPress is Perl via CGI and you just drop application into web directory to have it working immediately - same with updates - just overwrite files and Apache picks them up immediately.
In case of Rails is that you don't know target deployment architecture thus restarting application may not be trivial. E.g. with passenger I can just do touch tmp\restart.txt and then all instances are killed and started again. Some deployments may need init.d script restart invocation.
Maybe you could recommend or prepare a ready to use deployment model which supports autoupdate. In other cases users could do updates manually.

How to deploy a Rails app to Dreamhost?

I'm kind of lost, I try to deploy my application on a shared dreamhost server.
Now everything works fine locally. It's my first try at Rails, and I'm not really a programmer or sysadmin, just hacking something together.
On Dreamhost, if I start webrick, it works fine on port 3000, but webricks gets killed pretty quickly, I guess that makes sense.
So what do I have to do to make it run?
I enabled fastcgi support and mod_rails.
Now, how do I get the app constantly running?
I keep reading about having to do things to .htaccess and to dispatch.fcgi.
But I can't find any dispatch file in my rails app (2.3.2).
Do I have to create that one manually? Doesn't really feeld very rails-like to me. I didn't really manage to find out what this dispatch file does, and why it's needed.
Any help would be greatly appreciated.
I looked at the DH Wiki, but couldn't figure it out (http://wiki.dreamhost.com/Rails)
Additions:
I enabled mod_rails and pointed to the public directory (I had already done that).
I keep getting an error: screencast.com/t/KamqVawk
Hm, server logs look like there is actually no request, so this might be a problem on dreamhosts end. It's strange I see that the access.log show a new change date, when I try to access the page, but there is no request noted, error.log is empty too.
Dreamhost Support Answer:
The server was up to date, so that wasn't the problem.
They proposed freezing the Gems, which I did (see: http://wiki.dreamhost.com/Freezing_Gems)
But it didn't help.
I guess I'm giving up, and looking into hosting which is specialized for rails.
Thx for all your help!
Changed to hostingrails.com
I got the app working on hostingrails.com, passenger on hostingrails.com showed me errors, which weren't shown by dreamhost or mongrel. By correcting these errors, I got the app working.
Dreamhost won't let you use webrick if you're using shared hosting. You can either use FastCGI or Passenger to host Rails on shared DH (mongrel is an option if you upgrade to DreamhostPS, but that's obviously more expensive).
For FastCGI, you will need a dispatch.fcgi file (older versions of Rails would generate one when you created a new Rails app, but that stopped around 2.2 if I remember correctly) as well as code in your .htaccess to send requests to the dispatcher. See the Dreamhost Ruby on Rails wiki page for details about setting up FastCGI.
The more preferable option is to set up your application to run on Phusion Passenger (aka mod_rails). It should be pretty simple through your Dreamhost panel, you just need to enable the domain to use mod_rails, and then set the directory for the domain to the public directory of your application. See the Passenger wiki page for more details.

Resources