Publish to a Linux Production Environment

By Sourabh Shirhatti

In this guide, we will cover setting up a production-ready ASP.NET environment on an Ubuntu 14.04 Server.

We will take an existing ASP.NET Core application and place it behind a reverse-proxy server. We will then setup the reverse-proxy server to forward requests to our Kestrel web server.

Additionally we will ensure our web application runs on startup as a daemon and configure a process management tool to help restart our web application in the event of a crash to guarantee high availability.

Prerequisites

  1. Access to an Ubuntu 14.04 Server with a standard user account with sudo privilege.

  2. An existing ASP.NET Core application.

Copy over your app

Run dotnet publish from your dev environment to package your application into a self-contained directory that can run on your server.

Before we proceed, copy your ASP.NET Core application to your server using whatever tool (SCP, FTP, etc) integrates into your workflow. Try and run the app and navigate to http://<serveraddress>:<port> in your browser to see if the application runs fine on Linux. I recommend you have a working app before proceeding.

[!NOTE] You can use Yeoman to create a new ASP.NET Core application for a new project.

Configure a reverse proxy server

A reverse proxy is a common setup for serving dynamic web applications. The reverse proxy terminates the HTTP request and forwards it to the ASP.NET application.

Why use a reverse-proxy server?

Kestrel is great for serving dynamic content from ASP.NET, however the web serving parts aren’t as feature rich as full-featured servers like IIS, Apache or Nginx. A reverse proxy-server can allow you to offload work like serving static content, caching requests, compressing requests, and SSL termination from the HTTP server. The reverse proxy server may reside on a dedicated machine or may be deployed alongside an HTTP server.

For the purposes of this guide, we are going to use a single instance of Nginx that runs on the same server alongside your HTTP server. However, based on your requirements you may choose a different setup.

Install Nginx

sudo apt-get install nginx

[!NOTE] If you plan to install optional Nginx modules you may be required to build Nginx from source.

We are going to apt-get to install Nginx. The installer also creates a System V init script that runs Nginx as daemon on system startup. Since we just installed Nginx for the first time, we can explicitly start it by running

sudo service nginx start

At this point you should be able to navigate to your browser and see the default landing page for Nginx.

Configure Nginx

We will now configure Nginx as a reverse proxy to forward requests to our ASP.NET application

We will be modifying the /etc/nginx/sites-available/default, so open it up in your favorite text editor and replace the contents with the following.

server {
listen 80;
location / {
proxy_pass http://localhost:5000;
proxy_http_version 1.1;
proxy_set_header Upgrade $http_upgrade;
proxy_set_header Connection keep-alive;
proxy_set_header Host $host;
proxy_cache_bypass $http_upgrade;
}
}

This is one of the simplest configuration files for Nginx that forwards incoming public traffic on your port 80 to a port 5000that your web application will listen on.

Once you have completed making changes to your nginx configuration you can run sudo nginx -t to verify the syntax of your configuration files. If the configuration file test is successful you can ask nginx to pick up the changes by running sudo nginx -s reload.

Monitoring our Web Application

Nginx will forward requests to your Kestrel server, however unlike IIS on Windows, it does not mangage your Kestrel process. In this tutorial, we will use supervisor to start our application on system boot and restart our process in the event of a failure.

Installing supervisor

sudo apt-get install supervisor

[!NOTE] supervisor is a python based tool and you can acquire it through pip or easy_install instead.

Configuring supervisor

Supervisor works by creating child processes based on data in its configuration file. When a child process dies, supervisor is notified via the SIGCHILD signal and supervisor can react accordingly and restart your web application.

To have supervisor monitor our application, we will add a file to the /etc/supervisor/conf.d/ directory.

/etc/supervisor/conf.d/hellomvc.conf

[program:hellomvc]
command=/usr/bin/dotnet /var/aspnetcore/HelloMVC/HelloMVC.dll
directory=/var/aspnetcore/HelloMVC/
autostart=true
autorestart=true
stderr_logfile=/var/log/hellomvc.err.log
stdout_logfile=/var/log/hellomvc.out.log
environment=HOME=/var/www/,ASPNETCORE_ENVIRONMENT=Production
user=www-data
stopsignal=INT
stopasgroup=true
killasgroup=true

Once you are done editing the configuration file, restart the supervisord process to change the set of programs controlled by supervisord.

sudo service supervisor stop
sudo service supervisor start

Start our web application on startup

In our case, since we are using supervisor to manage our application, the application will be automatically started by supervisor. Supervisor uses a System V Init script to run as a daemon on system boot and will susbsequently launch your application. If you chose not to use supervisor or an equivalent tool, you will need to write a systemd or upstart or SysVinit script to start your application on startup.

Viewing logs

Supervisord logs messages about its own health and its subprocess' state changes to the activity log. The path to the activity log is configured via the logfile parameter in the configuration file.

sudo tail -f /var/log/supervisor/supervisord.log

You can redirect application logs (STDOUT and STERR) in the program section of your configuration file.

tail -f /var/log/hellomvc.out.log

Securing our application

Enable AppArmor

Linux Security Modules (LSM) is a framework that is part of the Linux kernel since Linux 2.6 that supports different implementations of security modules. AppArmor is a LSM that implements a Mandatory Access Control system which allows you to confine the program to a limited set of resources. Ensure AppArmor is enabled and properly configured.

Configuring our firewall

Close off all external ports that are not in use. Uncomplicated firewall (ufw) provides a frontend for iptables by providing a command-line interface for configuring the firewall. Verify that ufw is configured to allow traffic on any ports you need.

sudo apt-get install ufw
sudo ufw enable sudo ufw allow 80/tcp
sudo ufw allow 443/tcp

Securing Nginx

The default distribution of Nginx doesn't enable SSL. To enable all the security features we require, we will build from source.

Download the source and install the build dependencies

# Install the build dependencies
sudo apt-get update
sudo apt-get install build-essential zlib1g-dev libpcre3-dev libssl-dev libxslt1-dev libxml2-dev libgd2-xpm-dev libgeoip-dev libgoogle-perftools-dev libperl-dev # Download nginx 1.10.0 or latest
wget http://www.nginx.org/download/nginx-1.10.0.tar.gz
tar zxf nginx-1.10.0.tar.gz

Change the Nginx response name

Edit src/http/ngx_http_header_filter_module.c

static char ngx_http_server_string[] = "Server: Your Web Server" CRLF;
static char ngx_http_server_full_string[] = "Server: Your Web Server" CRLF;

Configure the options and build

The PCRE library is required for regular expressions. Regular expressions are used in the location directive for the ngx_http_rewrite_module. The http_ssl_module adds HTTPS protocol support.

Consider using a web application firewall like ModSecurity to harden your application.

./configure
--with-pcre=../pcre-8.38
--with-zlib=../zlib-1.2.8
--with-http_ssl_module
--with-stream
--with-mail=dynamic

Configure SSL

  • Configure your server to listen to HTTPS traffic on port 443 by specifying a valid certificate issued by a trusted Certificate Authority (CA).

  • Harden your security by employing some of the practices suggested below like choosing a stronger cipher and redirecting all traffic over HTTP to HTTPS.

  • Adding an HTTP Strict-Transport-Security (HSTS) header ensures all subsequent requests made by the client are over HTTPS only.

  • Do not add the Strict-Transport-Security header or chose an appropriate max-age if you plan to disable SSL in the future.

Add /etc/nginx/proxy.conf configuration file.

[!code-nginxMain]

Edit /etc/nginx/nginx.conf configuration file. The example contains both http and server sections in one configuration file.

[!code-nginxMain]

Publish to a Linux Production Environment的更多相关文章

  1. 【asp.net core】Publish to a Linux-Ubuntu 14.04 Server Production Environment

    Submary 又升级了,目录结构有变化了 . project.json and Visual Studio 2015 with .NET Core On March 7, 2017, the .NE ...

  2. [转]The Production Environment at Google (part 2)

    How the production environment at Google fits together for networking, monitoring and finishing with ...

  3. [转]The Production Environment at Google

    A brief tour of some of the important components of a Google Datacenter.   A photo of the interior o ...

  4. Implement a deployment tool such as Ansible, Chef, Puppet, or Salt to automate deployment and management of the production environment

    Implement a deployment tool such as Ansible, Chef, Puppet, or Salt to automate deployment and manage ...

  5. EF中三大开发模式之DB First,Model First,Code First以及在Production Environment中的抉择

    一:ef中的三种开发方式 1. db first... db放在第一位,在我们开发之前必须要有完整的database,实际开发中用到最多的... <1> DBset集合的单复数... db ...

  6. 在Linux上以服务的方式运行ASP.NET Core站点

    更新:用supervisor是更好的解决方法,详见 Linux下为 dotnet 创建守护进程 要在生成环境下在Linux服务器上跑ASP.NET Core站点,首先要解决的问题是以服务的方式运行AS ...

  7. 发布ASP.NET Core程序到Linux生产环境

    原文翻译:Publish to a Linux Production Environment 作者:Sourabh Shirhatti 在这篇文章里我们将介绍如何在 Ubuntu 14.04 Serv ...

  8. 怎么部署 .NET Core Web项目 到linux

    .NET Core is free, open source, cross platform and runs basically everywhere. STEP 0 - GET A CHEAP H ...

  9. ASP.NET Core 中文文档 第三章 原理(12)托管

    原文:Hosting 作者:Steve Smith 翻译:娄宇(Lyrics) 校对:何镇汐.许登洋(Seay) 为了运行 ASP.NET Core 应用程序,你需要使用 WebHostBuilder ...

随机推荐

  1. Asp.netMVC中Html.Partial,RenderPartial,Action,RenderAction区别和用法

    https://www.2cto.com/kf/201702/602449.html Partial 和RenderPartial: 这两个的性质都是一样, 只指把一个个View给镶入进来, 只是回传 ...

  2. SEM(搜索引擎营销)

    ylbtech-Miscellaneos:SEM(搜索引擎营销) 搜索引擎营销:英文Search Engine Marketing ,我们通常简称为“SEM”.就是根据用户使用搜索引擎的方式利用用户检 ...

  3. LeetCode 114| Flatten Binary Tree to Linked List(二叉树转化成链表)

    题目 给定一个二叉树,原地将它展开为链表. 例如,给定二叉树 1 / \ 2 5 / \ \ 3 4 6 将其展开为: 1 \ 2 \ 3 \ 4 \ 5 \ 6 解析 通过递归实现:可以用先序遍历, ...

  4. 自动化打包 Jenkins 持续集成 Git Gradle MD

    Markdown版本笔记 我的GitHub首页 我的博客 我的微信 我的邮箱 MyAndroidBlogs baiqiantao baiqiantao bqt20094 baiqiantao@sina ...

  5. JS条件判断

    JavaScript 是一种可以在浏览器中运行的脚本语言,是一种弱语言(相对于C,C#,JAVA而言),只要是计算机语言就会使用到条件判断式,而JavaScript作为一种“弱”语言,它的条件判断常常 ...

  6. delphi判断.net FrameWork是否已安装

    判断系统中.NET FrameWork已安装版本的方法很多,比如检查C:\Windows\Microsoft.NET\Framework\下的子目录,但是子目录往往是包含版本号,比如v2.0.5072 ...

  7. 原生JS实现new方法、new一个对象发生的四部、new里面常用的优先级

    一.js中new一个对象的过程 首先了解new做了什么,使用new关键字调用函数(new ClassA(…))的具体步骤: 1.创建一个新对象: var obj = {}; 2.设置新对象的const ...

  8. 【cocos2dx中Node类getParent和getChildByTag()】学习体会

    參考http://cn.cocos2d-x.org/doc/cocos2d-x-3.0/d3/d82/classcocos2d_1_1_node.html 当中和child.parent有关的成员函数 ...

  9. spark0.9分布式安装

    http://blog.csdn.net/myboyliu2007/article/details/18990277 spark安装包:spark-0.9.0-incubating-bin-hadoo ...

  10. JPA(六):映射关联关系------映射单向一对多的关联关系

    映射单向一对多的关联关系 新建项目项目请参考<JPA(二):HellWord工程>,基于上一章讲解的<JPA(五):映射关联关系------映射单向多对一的关联关系>中的例子进 ...