How would you deploy a Kelp app?

[Brad]

I've created a nifty little API on my dev machine (which did not take long thanks to Kelp). I now need to deploy it to a staging VM for proper testing, and to allow the customer to try it out. I'm just wondering how you guys deploy your Kelp apps?

Cheers,
Brad

[Stefan Geneshky]

You would deploy a Kelp app the same way you would deploy a Plack app, so feel free to reach out to the Plack community as well.

This is how I usually deploy my projects:

start_server --port=8080 -- plackup -s Starman -E deployment app.psgi

start_server is provided by Server::Starter - http://search.cpan.org/~kazuho/Server-Starter-0.17/lib/Server/Starter.pm
It allows you to gracefully restart your application by sending the process a HUP signal.

I use Starman, but you can just as successfully use Starlet, depending on your needs.
It is important that you specify -E deployment, because this instructs Kelp to you your deployment.pl config file.

Alternatively,

​ instead of plackup,​

you can use uWSGI, which is written in C and it's very fast.
uwsgi --plugins psgi --init app.ini

Your app.ini could look something like:
[uwsgi]
http-socket = :8080
psgi = app.psgi
disable-logging = True

​You will have to install uwsgi yourself. I actually had to download its source and compile it.​

The above both methods start the application at port 8080. I use nginx, so I add to its configuration something in the lines of:

upstream ifnx {
        server 127.0.0.1:8080;
}

server {
        listen 80;
        server_name ifnx.com www.ifnx.com;
        client_max_body_size 100M;

        root /usr/share/nginx/html;
        access_log /var/log/nginx/ifnx_access.log;
        error_log /var/log/nginx/ifnx_error.log;

        location / {
                proxy_read_timeout 300;
                proxy_pass http://ifnx;
                proxy_set_header Host $host;
                proxy_set_header X-Forwarded-For $proxy_add_x_forwarded_for;
                proxy_set_header X-Forwarded-HTTPS 0;
                proxy_set_header X-Real-IP $remote_addr;
        }
}

[Maurice Aubrey]

On Fri, Sep 26, 2014 at 3:36 AM, Brad <br...@perlpowered.com> wrote:

I've created a nifty little API on my dev machine (which did not take long thanks to Kelp). I now need to deploy it to a staging VM for proper testing, and to allow the customer to try it out. I'm just wondering how you guys deploy your Kelp apps?

For Starman, I use this init script in production:

https://github.com/mla/starman-init

[Brad]

Ah nice, I have never heard of uwsgi.. will have to try it out! Yes, I generally use the Starman/nginx method, then supervise daemon tools to start and stop the service. It also monitors the process and restarts it if it gets unexpectedly killed, or machine rebooted - http://cr.yp.to/daemontools.html