RSS

Category Archives: Fedora + CentOs

Các thông tin và hướng dẫn về Fedora nói riêng và các dòng RedHat nói chung

Apache Web Server Tutorial for Linux


Introduction

The Apache Web Server is one of the most popular open-source HTTP Servers that exists today. It’s a powerful, secure, and fully featured web server that can be found hosting anything from personal web sites to corporate domains. If you have a DSL or Cable connection with a static IP then you can use the Apache web server to host your web site (as long as your ISP doesn’t have a problem with you running a server on their network). This tutorial will teach you how to set up your own Apache web server on Linux so you can host your own web sites either at home or work. You’ll learn how to install the Apache HTP Server with SSL and PHP support, use chkconfig to set Apache to auto-start on boot, edit the Apache httpd.conf configuration file for your server settings, create a directory structure for your web site, create the VirtualHost configuration file for your web site, start the Apache service, verify your site is working, go over the init script options, and finally show you some additional options and settings for Apache such as graceful restarts, htaccess usage, user name and password directory protection, and custom error pages. The two things I won’t be covering are setting up DNS for your domain and configuring your firewall to open ports 80 and 443 for HTTP and SSL respectively, however you can read my Bind DNS tutorial and iptables firewall tutorial for information on how to do it. I’ll be using CentOS 4.2 and Apache 2.x will be installed using yum instead of source. Let’s get started.

Installing Apache HTTP Web Server

We’ll be using yum to download and install the Apache RPM’s from the CentOS repository. Apache shouldn’t be loaded on your server unless you selected it during the operating system installation process. If you know you have Apache installed then you can skip this section, but if it’s not or you’re not sure then you should perform the following tests to see if the Apache web server is on your system. To do this we’ll use the which command to look for the httpd binary in the environment variable PATH.

which httpd

[graphical representation of executing 'which httpd']

If you see something similar to the above image then the Apache binary doesn’t seem to be found on the server. Our next step is to install Apache by using yum. Issue the next command.

yum install httpd

[graphical representation of executing 'yum install httpd']

Two RPM packages will be downloaded from the CentOS repository (httpd-2.0.52-22.ent.centos4.i386.rpm and httpd-suexec-2.0.52-22.ent.centos4.i386.rpm) and installed automatically. If everything goes well you should now have the Apache web server on your system. You might also want to load the Apache documentation so you’ll have the man pages available. This is optional but highly recommended. Use yum again and download the Apache manual (yum install httpd-manual). Next we’ll use yum to install mod_ssl for Secure Sockets Layer (SSL) support.

Installing mod_ssl for Secure Sockets Layer (SSL) Support

Secure Sockets Layer, also known as SSL for short, provides encrypted commnications between hosts. SSL can be found in e-commerce web sites and any other site that requires sensitive information being transmitted protected from eavesdropping. In a nutshell, the hosts negotiate a protocol they both can understand and then switch to secure communications. The data passed back and forth between the hosts is encrypted with a 128-bit key (hence 128-bit SSL). In a typical SSL setup, the web server has SSL support enabled and a certificate that verifies the company’s identity. These certificates are initially generated as Certificate Signing Requests (CSR) by the server’s administrator. The CSR is then signed by a third party, a Certificate Authority (CA), that validates the company’s identity and makes the CSR into a full certificate. There are many CA’s like VeriSignThawteGeoTrustInstantSSL (aka The Comodo Group), FreeSSL, and many others that sign CSR’s. Prices vary from company to company, but as long as they support the mainstream browsers you shouldn’t run into any problems. If you’d like to learn more about Secure Sockets Layer then you should check out the Wikipedia Transport Layer Security page.

You may be wondering where does mod_ssl fits into all of this? On Linux and Unix distributions, there is a software package called OpenSSL that performs all the cryptography and mod_ssl is the Apache interface to OpenSSL. It’s mod_ssl that makes it possible for Apache to have SSL capability. We’ll now install the mod_ssl RPM (mod_ssl-2.0.52-22.ent.centos4.i386.rpm) for Apache using yum.

yum install mod_ssl

[graphical representation of executing 'yum install mod_ssl']

mod_ssl should now be installed. If you didnt have OpenSSL installed then it should have been picked up as a dependency and was installed as well. We have one more package to install and that’s the PHP scripting language.

Installing PHP for Dynamic Web Pages

PHP is a scripting language for creating dynamic web pages. PHP allows you to include code in your web pages to be processed server-side and the resulting HTML is sent to the user’s web browser. With PHP you can add database access, read and write files to the server’s filesystem, generate dynamic graphics like graphs and security image codes, and create security control mechanisms for fencing off parts of your web site. PHP has many other uses and that’s what makes it a powerful tool for creating dynamic web pages.

If you’re not a programmer or will only he hosting a static HTML web pages, then PHP won’t be of any use to you. However, if you want to learn PHP then you should install the PHP modules for Apache. We’ll use yum one last time to install PHP on our server. There are many PHP RPM modules, one being the actually binary and modules for Apache and the rest are support features for MySQL, XML, ODBC, and others to tie those subsystems into PHP. It’s a good idea to install all the PHP modules since you never know what you’ll need now or in the future. Now we’ll use yum to install PHP.

yum install php*

[graphical representation of executing 'yum install php*']

As you can see above, we used a wildcard to tell yum to install anything starting with the word ‘php’. On my server yum downloaded and installed fifteen packages, however yours may be different depending on its configuration. Now that PHP is installed we can move on to setting the Apache service to automatically start on bootup or in the event of a reboot.

Setting Apache to Start on Bootup with chkconfig

The Apache web server will need to be set to autostart when the server boots. The Apache RPM didn’t set this up for us so we’ll have to do it ourselves using chkconfig (Note: you can also use setup to turn on the Apache service). We’re going to have Apache start on run levels 2, 3, and 5.

chkconfig –level 235 httpd on
chkconfig –list httpd

[graphical representation of executing 'chkconfig --level 235 httpd on']

The second chkconfig command lists the run levels Apache is configured to start on. If you don’t have X Windows installed then you may want to omit run level 5 (Multi-User Mode – boot up in X Windows). To learn more about Linux Run Levels you may want to check out this page on NetworkClue. Now that Apache is installed and set to start up we’ll move on to configuring the Apache web server by editing httpd.conf, the Apache server configuration file.

Configuring Apache Server Settings (httpd.conf)

Apache’s main configuration file is called httpd.conf and is located in /etc/httpd/conf/. The default httpd.conf will work without any changes, however we want to customize Apache a little bit. Our main focus is to setup the use of Virtual Hosts so we can run as many web sites as we want using a single IP address. Also, we want to simplify the management of our Virtual Hosts without cluttering httpd.conf with our entries. Before we get to any of that we will first configure our Apache web server.

I will make the assumption that you have a single WAN based IP address. For this example I’m using 192.168.1.210 with a hostname of node2.centos (yes, it’s a LAN IP but pretend it’s WAN). You can use your IP or hostname from /etc/hosts, but I suggest using your hostname because if you ever change IP’s all you have to do is update /etc/hosts with the new address. If you do use an IP then you’ll have to change all instances of it in httpd.conf. If you don’t know what your IP address is or never set up /etc/hosts, you can find your address by using ifconfig (look at the number to the right of ‘inet addr:’).

We’ll start with opening httpd.conf in the nano text editor, but before that you should make a copy of your existing httpd.conf file. If you encounter any problems you’ll still have the original to fall back to.

cd /etc/httpd/conf
cp httpd.conf httpd.conf.old
nano httpd.conf

Apache’s httpd.conf is filled with many helpful comments to tell you what each configuration directive does. Scroll down to line 133 as shown in the picture below. FYI, if you ever want to know what line number you’re on in nano, press CTRL-C and nano will show you. We’re looking for the line that says ‘Listen 80’. We won’t be changing this directive since we want Apache listen on all IP addresses set up on the server. I wanted to show you where to change it if you needed to bind Apache to only one address.

[graphical representation of executing 'cd /etc/httpd/conf','cp httpd.conf httpd.conf.old','nano httpd.conf']

Move on down to line number 235 where it says ‘ServerAdmin root@localhost’. The Apache ServerAdmin directive is for the administrator’s email address of the server. End users encountering any problems with the server would use this email address to notify the sysadmin. Comment out the existing ServerAdmin line with a pound symbol and enter a new line below with your email address. For this example I used admin@node2.centos.

#ServerAdmin root@localhost
ServerAdmin admin@your-domain.com

[graphical representation of adding 'ServerAdmin admin@your-domain.com' to httpd.conf]

Below the ServerAdmin directive is ServerName. The Apache ServerName directive is for Apache to identify itself which is typically the hostname of the server. You’ll want to specify your hostname or IP address instead of a valid DNS name especially if you’re hosting many web sites. There are some people that may disagree with this method but I think it’s better to keep all the generic information in httpd.conf and use specifics in the VirtualHost configuration files. Add a new blank line below the commented ServerName directive and add your hostname or IP. For this server I used my hostname node2.centos.

#ServerName new.host.name:80
ServerName yourhostname

[graphical representation of adding 'ServerName yourhostname' to httpd.conf]

The next directive we’ll edit is very close to the end of the file. Scroll all the way down to line number 1005 (tip: use the Page Down key on your keyboard) and look for ‘#NameVirtualHost *:80’. The NameVirtualHost directive tells Apache that we want to use name-based virtual hosting, or in simpler terms, a bunch of web sites all using the same IP address. Virtual hosting is made possible because when a web browser goes to a site, for example http://www.xenocafe.com, the web site address is passed as part of the HTTP header (Host: http://www.xenocafe.com). This allows Apache to distinguish between different hosts sharing the same IP address. Add a couple blank lines underneath ‘#NameVirtualHost *:80’ and on the line right below put NameVirtualHost and your hostname or IP address followed by a :80. The ‘:80’ means we’re using name-based virtual hosting for the HTTP protocol. SSL will be configured through VirtualHost configuration files due to its nature. You’ll learn why when we create our virtual hosts. For this server I used my hostname node2.centos.

#NameVirtualHost *:80
NameVirtualHost yourhostname:80

[graphical representation of adding 'NameVirtualHost yourhostname:80' to httpd.conf]

We’re almost done. The last thing we need to do is create a default virtual host to respond to requests when someone vists our IP address and not our domain name. You may or may not want to add this virtual host and it’s totally up to you, but I personally don’t want anyone going directly to my IP address. I prefer they visit my web site by name only. A VirtualHost entry can have many directives which I’ll explain later, but what you need to know for now is the example I provide responds by IP address visits. At the very end of the file there is a line that instructs Apache to load any configuration files found in the /etc/httpd/conf/vhosts/ directory (which we’ll create later). This is my way of keeping the virtual hosts separate from httpd.conf and most importantly, making hosts easier to manage.

<VirtualHost yourhostname:80> ServerAdmin you@your-domain.com ServerName your_ip_address DocumentRoot /www ErrorLog logs/error_log CustomLog logs/access_log combined </VirtualHost> # include VirtualHosts config files Include conf/vhosts/*.conf

[graphical representation of adding an IP based virtual host to httpd.conf and 'Include conf/vhosts/*.conf']

We’re done editing httpd.conf. Let’s save our changes (CTRL-O) and exit nano (CTRL-X). Now we’ll move on to creating a web site directory structure and user account for SSH access and S/FTP file uploads.

Creating the Web Site Directory Structure

Our web site directory schema should be simple yet structured. The common approach would be to use /home as the root but we won’t be doing that. Our web site parent root will be /www and from within it create a directory for each domain we’ll be hosting. Within each domain directory there will be a set of common directories (html, html/cgi-bin, databases, and logs). Reading what I wrote doesn’t make much sense so here it is visually. For this example and the rest of the tutorial, we’ll say we acquired the domain your-domain.com and will be configuring our server for it.

/www (root for all hosted domains) /www/your-domain (domain directory) /www/your-domain/html (directory for your web site files) /www/your-domain/html/cgi-bin (CGI directory for executing Perl scripts) /www/your-domain/databases (databases for this web site stored on a per site basis) /www/your-domain/logs (web site access and error logs are stored here)

The databases directory is optional. If you read my mapping mysql databases tutorial then you can migrate your databases outside the default /var/lib/mysql into the databases directory. This will allow you to store your databases on a per site basis in the web site’s home directory.

Creating the Directory Root (one-time only)

We’ll start creating our web site directory structure by making /www. Creating this directory is a one-time process and will be home to all our domains we’re hosting (any domains we add will be stored in the /www directory).

mkdir /www
chown root.root /www

[graphical representation of executing 'mkdir /www','chown root.root /www']

After we create the directory we chown it to uid:root gid:root.

Creating Web Site Domain Directories

For each web site you’re hosting you’ll create a directory structure within the parent /www. The directory should be named the same as the domain name with or without the TLD (.com, .net, etc). I’ll let you decide which way you want your directories to be named, but I tend to prefer them without the TLD. Now we’ll create our domain’s directories.

cd /www
mkdir your-domain
cd your-domain
mkdir html
mkdir html/cgi-bin
mkdir databases
mkdir logs

[graphical representation of executing 'cd /www','mkdir your-domain','cd your-domain','mkdir html','mkdir html/cgi-bin','mkdir darabases','mkdir logs']

Now that the directory structure is created we’ll need to add a system user that owns the directory base and can log in to upload files.

Adding a Linux User Account for the Web Site

The user account will be able to FTP or Secure FTP (SFTP) into the server to upload and download web site content files like HTML, CGI scripts, and PHP pages. This user account will have ownership of the domain’s directory base and be locked into it. The user name we’ll assign will be the same as the domain name (your-domain) to make it easier to distinguish between it and the other Linux accounts, however you can use any name you want for your account. One thing to keep in mind is when you create this account you’ll have to make a choice whether this account will be granted SSH access. If you want to disallow SSH access then you’ll need to change the shell to /sbin/nologin. SSH login attempts will fail when you use /sbin/nologin as the account’s shell (valid login credentials will be accepted but the connection will be terminated). To create the web site system account, use one of the following commands.

useradd -d /www/your-domain/ domain_user
or
useradd -d /www/your-domain/ -s /sbin/nologin domain_user

[graphical representation of executing 'useradd -d /www/your-domain/ -s /sbin/nologin/ your-domain','cat /etc/passwd | grep -i your-domain']

The first command will set the account’s home directory (-d path) to /www/your-domain and creates a new system account with the user name of domain_user (last parameter of the command). The shell will be set to the default, which is typically bash unless you changed it. The second command is identical to the first except that the shell (-s shell) is set to /sbin/nologin to prevent SSH logins.

After the account has been created you’ll need to set a password using the passwd command.

passwd domain_user
[enter password]
[confirm password]

[graphical representation of executing 'passwd domain_user']

Now that the system account is ready we’ll move on to changing the directory base ownership.

Changing Web Site Directory Ownership & Permissions

We’ve created the directories, added the account, the only thing left to do is change the ownership of the directories to the new user account. Right now the directories are owned by root and if you were to try using FTP to upload some files you’d get the dreaded Permission Denied error. To change ownership of the domain directory and all its subdirectories we’ll use the chown command with the -R flag for recursive.

cd /www
chown -R domain_user:domain_user your-domain

[graphical representation of executing 'cd /www','chown -R domain_user:domain_user your-domain']

If you created the databases directory then you’ll need to change the owner to the mysql user otherwise MySQL won’t be able to write to the database files. All directories, subdirectories, and files starting from databases should be owned by the mysql user (or whatever user account you have the MySQL server running as).

cd your-domain
chown -R mysql:mysql databases

[graphical representation of executing 'cd your-domain','chown -R mysql:mysql databases','ls -la']

That’s it for creating our directory structure. If you have more domains to add then repeat the steps we went through except for the first one of creating the /www directory root. Next we’ll create the VirtualHost configuration files to let Apache know the details of our web site.

Creating VirtualHost Configuration Files

A VirtualHost configuration file tells Apache the specifics about your web site. It contains information like what IP and port number your site runs on, the server name and any aliases, the local filesystem path (or document root) of your web pages, location of your cgi-bin, file names and paths of your access and error logs, custom error documents for HTTP errors like 404 File Not Found, and can have many other items. For every web site you want Apache to host you must create a virtual host configuration. You’ve already gotten a taste of a virtual host config when we added the default VirtualHost for handling IP address requests in httpd.conf, so what we’re going to be doing now is nothing new. I’m now going to show you an easy way of managing all your virtual hosts.

The VirtualHost Configuration File Directory (/etc/httpd/conf/vhosts)

When I first learned Linux from a friend some years ago, he showed me how to configure Apache so I could set up my web server. I looked at his httpd.conf and he had all his virtual hosts in there, so I figured there was nothing unusual about that as I’m sure there are many people that did the same thing. He hosted a lot of domains for friends, his business, and his own stuff, and I noticed that trying to locate a web site in his massive httpd.conf file was nothing short of a headache. Although I didn’t doubt his knowledge of Apache I kept thinking there had to be an easier way of managing virtual hosts. Through reading the online Apache Web Server documentation I came across the Include directive.

The Include directive, much like in C/C++, PHP, and other languages, allows you to insert the contents of another file into the calling file. However, in the case of Apache, the contents of a single file or of several files located in a directory that match the Include statement would be included. If you remember when we were editing httpd.conf, at the very end we added the line ‘Include conf/vhosts/*.conf’. That line tells Apache to look in /etc/httpd/conf/vhosts for any *.conf files and include their contents as part of httpd.conf. This allows us to separate out our VirtualHost definitions as their own unique files, on a per domain basis, from the main httpd.conf file. This makes administration easy for several reasons. Finding a virtual host is easy because all you do is look in the vhosts directory for its file, disabling a domain is as simple as renaming the file so it doesn’t end in .conf, and adding new domains is accomplished by creating a new virtual host file (copy an exisiting file and edit the changes). Include makes virtual host management a piece of cake.

We’ll now create the /etc/httpd/conf/vhosts directory (this is a one-time process).

cd /etc/httpd/conf
mkdir vhosts
cd vhosts

[graphical representation of executing 'cd /etc/httpd/conf','mkdir vhosts','cd vhosts']

We can start creating VirtualHost configuration files now. I’ll show you how to create one for the HTTP protocol as well as one for Secure Sockets Layer (SSL).

Creating a HTTP Virtual Host

When creating virtual host config files we should use a good naming convention that’s as descriptive as possible. What’s worked for me is the format of host.domain.tld.port.conf or in simple terms your-domain.com.80.conf. You may have noticed I excluded host from the file name. When you visit a domain such as redhat.com, or http://www.redhat.com, they most likely take you to the same web site, so www as a host can be excluded. The only time I use host as part of the file name is when I am adding a subdomain such as forum.example.com or blog.example.com and so forth. Let’s create the HTTP virtual host file.

nano your-domain.com.80.conf

Type in or copy/paste the configuration data below. Be sure to replace yourhostname with your system’s hostname or IP address, ServerAdmin with your email address, ServerName and ServerAlias with your real domain name, and your-domain with the directory name you created earlier for your web site. When you’re done, save your changes (CTRL-O) and exit nano (CTRL-X).

<VirtualHost yourhostname:80> ServerAdmin admin@your-domain.com ServerName your-domain.com ServerAlias www.your-domain.com DocumentRoot /www/your-domain/html ScriptAlias /cgi-bin/ /www/your-domain/html/cgi-bin/ ErrorLog /www/your-domain/logs/error_log CustomLog /www/your-domain/logs/access_log combined </VirtualHost>

[graphical representation of a name-based virtual host for http]

I’ll clarify what each line does in the virtual host configuration file between the <VirtualHost></VirtualHost> tags.

  • ServerAdmin is used to specify the server administrator’s email address. User’s will see this address on server generated error pages.
  • ServerName is the name of the server or DNS name. Apache will match this name against the HTTP host header sent by the user’s browser. If it’s a match then this is the configuration data used to process the request.
  • ServerAlias is identical to ServerName except ServerName is used once and ServerAlias can be used many times. If your web site will respond to more than one DNS name, make sure you use one ServerName directive and all other names are specified using ServerAlias. DNS names can be separated by commas for a single ServerAlias directive (ServerAlias dns_name1,dns_name2,dns_name3) or you can use many ServerAlias directives for each DNS name placed on a line each their own.
  • DocumentRoot defines the file system path to where the web page files are located. This is the same directory where you will place your HTML, PHP, and image files. For our setup, this is the /www/your-domain/html directory.
  • ScriptAlias defines the directory for executing cgi scripts, commonly known as cgi-bin. It takes two parameters with the first being the cgi-bin directory from your web site root path and the second being the full physical directory path on the filesystem. You can have as many cgi-cin directories as you need and they don’t need to be called cgi-bin.
  • ErrorLog defines the file to log any errors Apache encounters. Errors can include PHP script failures and files that don’t exist.
  • CustomLog defines the file to log any requests to the server. Requests include accessing pages, images, and other files like robots.txt. The second parameter sets the log entry format. We use the combined format since it is used by many Web Server Log Analysis scripts like Awstats.

You should read the online Apache 2.0 documentation for more information on these directives as well as other Apache web server features.

Creating a SSL Virtual Host

When we installed mod_ssl, a file called ssl.conf was written to /etc/httpd/conf.d that contains basic SSL configuration information. We’re going to use this as our template for adding a SSL based virtual host. What we’ll do is make a copy of it to our vhosts directory and remove the non-website specific information since it should only be loaded once by Apache. This non-website specific information pertains to loading the mod_ssl Apache module among other things while everything else in the file relates to configuring the SSL web site (what we’re interested in).

Let’s make a copy of ssl.conf to the vhosts directory which you should still be in (cd /etc/httpd/conf/vhosts). Following the same format for our virtual host config files (host.domain.tld.port.conf) as before, we will be naming our SSL virtual host file as your-domain.com.443.conf. The port number for SSL is 443 so we’ll use that as opposed to port 80 when we created the HTTP VirtualHost. After that we’l open the config file in nano so we can begin editing it.

cp /etc/httpd/conf.d/ssl.conf ./your-domain.com.443.conf
nano your-domain.com.443.conf

[graphical representation of executing 'cp /etc/httpd/conf.d/ssl.conf ./your-domain.com.443.conf','nano your-domain.com.443.conf']

Now comes the fun part. We need to delete a good portion of this file starting from the top and all the way down to the part that says ‘### SSL Virtual Host Context’ (roughly 83 lines total). Instead of using the DEL key, you can use CTRL-Kto cut the text line by line in nano. Starting from the top of the file, remove the lines of text until you reach the blank line above ‘### SSL Virtual Host Context’. Your file should look like the image below when you’re done.

[graphical representation of editing ssl virtual host config file]

Next we’ll edit the default web site specific directives to mold them to our web site. They’re actually the same directives we used when creating our HTTP virtual host, so you can see how easy this will be. I should point out that we have been using our hostname for the VirtualHost tag, but in the case of SSL I’d advise against unless you want to set up each host and their binding IP address in /etc/hosts (it’s a good idea if you have many IP’s). Secure Sockets Layer doesn’t work at the Application layer of the OSI model like HTTP does, so the Host part of the HTTP header doesn’t apply. So in regards to name-based virtual hosting with SSL, it’s non-existent. Replace ‘_default_’ with your IP address when editing your-domain.com.443.conf. The same rules apply as before with ServerName, ServerAlias, and the your-domain directory.

Here is a stripped down version of our SSL VirtualHost file minus the comments.

<VirtualHost yourip:443> DocumentRoot "/www/your-domain/html" ServerName your-domain.com ServerAlias www.your-domain.com ScriptAlias /cgi-bin/ /www/your-domain/html/cgi-bin/ ErrorLog logs/ssl_error_log TransferLog logs/ssl_access_log LogLevel warn SSLEngine on SSLCipherSuite ALL:!ADH:!EXPORT56:RC4+RSA:+HIGH:+MEDIUM:+LOW:+SSLv2:+EXP SSLCertificateFile /etc/httpd/conf/ssl.crt/server.crt SSLCertificateKeyFile /etc/httpd/conf/ssl.key/server.key <Files ~ "\.(cgi|shtml|phtml|php3?)$"> SSLOptions +StdEnvVars </Files> <Directory "/var/www/cgi-bin"> SSLOptions +StdEnvVars </Directory> SetEnvIf User-Agent ".*MSIE.*" \ nokeepalive ssl-unclean-shutdown \ downgrade-1.0 force-response-1.0 CustomLog logs/ssl_request_log \ "%t %h %{SSL_PROTOCOL}x %{SSL_CIPHER}x \"%r\" %b" </VirtualHost>

[graphical representation of revised ssl virtual host config file]

This configuration uses mod_ssl’s default SSL certificate and private key for encrypted communications. It’s not signed by a trusted authority nor does it have your personal or company credentials. Your web browser will complain about the certificate and throw up a dialog box when you visit your web site (unless you instruct the browser to save the certificate). If you’re thinking about doing e-commerce then you’ll need to invest in a real SSL certificate because nothing kills a customer’s confidence level like a warning screen while making a purchase online. What you would do is generate your Certificate Signing Request (CSR) on your server and then send the data to a Certificate Authority (CA) like Verisign. Once you pay the fees and prove your identity, they’ll sign your certificate to make it legitimate and then you can replace the values of SSLCertificateFile and SSLCertificateKeyFile with your files (in the case of Comodo SSL, you’d have to set the SSLCACertificateFile directive with their Bundle Certificate file as well). On the other hand, if you’re setting up SSL for your own personal use or for an intranet at work, self-signed SSL certificates work just fine. There is no difference between legitimate and self-signed certs except for the CA taking your money and stamping it with their sign of approval. The encryption protection is the same either way.

Save your changes (CTRL-O) to your-domain.com.443.conf and exit nano (CTRL-X). I don’t have a tutorial up for generating SSL certificates using OpenSSL, but I will soon. You can check Google for a site that details those instructions or wait for mine to be posted. Also, you should read up on the Apache SSL/TLS Encryption documention for more information about the directives and virtual host configurations for Apache SSL.

Starting Apache and Viewing the Web Site

Apache is set up, our directories have been created, and the VirtualHost config files are done, now is the time to start the Apache web server and see if everything worked. It’s taken us a long time to get here and if all goes well then Apache should start without any errors and your HTTP and SSL web sites should be available.

service httpd start

[graphical representation of executing 'service httpd start','nmap localhost']

If Apache failed to start for you, you should double check httpd.conf, your directories, and your virtual host configs for any typos or syntax mistakes. Apache can be cryptic at times when errors occur. You can verify virtual host files usinghttpd -S (or for older an older Apache version use httpd -t -D DUMP_VHOSTS). In the next section I’ll show you some tips for tracking down Apache errors.

httpd -S

[graphical representation of executing 'httpd -S']

With the Apache web server running you should have already opened ports 80 and 443 on your firewall so you can connect to Apache with a web browser to view your web site. Open up your favorite browser and navigate to your site. You should get the default CentOS Apache 2 Test Page unless you have already logged in and uploaded your content via S/FTP.

[graphical representation of visiting http virtual host in FireFox]

Standard HTTP works so how about our SSL version of the web site? Change the http:// to https:// in the address bar of your browser. You should be prompted by the security dialog I mentioned before. This is just a warning about your self-signed SSL certificate not being trusted by your browser.

[graphical representation of FireFox warning about self-signed certificate]

Click OK in FireFox or Yes in Internet Explorer to continue to the secure web site.

[graphical representation of visiting ssl virtual host in FireFox]

As you can see up in the FireFox address bar, there is a little yellow padlock icon that informs us that we are on a secure web site. All communications with this site will be 128-bit encrypted. At this time you may want to log in to your web site via Secure FTP (SFTP) or regular FTP to upload your web site files to the html directory using the system account you created earlier. You’re done! You’ve successfully configured the Apache web server on Linux.

Other Things to Know for Using Apache

Here are some common questions that users new to the Apache HTTP Server may ask (well, the questions I could think of at the time).

How can I restart Apache so that it sees the changes I made to httpd.conf and/or virtual hosts?
The Apache init script (/etc/init.d/httpd) accepts a number of options, four of the important ones being start, stop, restart, and graceful.

  • start tells the init script to start the Apache process and spawns the child processes.
  • stop tells the init script to terminate the Apache’s running process and in turn kills off the child processes. Any existing connections to the server will be terminated.
  • restart tells the init script to stop the Apache process and start it back up. Any changes to httpd.conf or virtual hosts will be read back in. Any existing connections will be terminated, so if users are in the middle of a download the transfers will stop.
  • graceful tells the init script to alert Apache to reload httpd.conf and virtual hosts. Existing connections will not be terminated but changes won’t be noticed until a new connection is established (pulling up another page or file from the server). Requests are handled by Apache child workers.

There are two ways of using these options. You can use the init script or the service command. To use the init script, enter this on the prompt /etc/init.d/httpd option, such as to restart the Apache server use /etc/init.d/httpd restart from the command line. To use the service command, which is what I showed when we started the Apache service, it’s service httpd option, such as to do a graceful restart use service httpd graceful from the command line.

Apache won’t start. It keeps saying FAILED when I try and start the service. What can I do to fix it?
This is probably one of the most difficult tasks to do since sometimes Apache doesn’t give an error or it’s too cryptic when it does. Errors usually occur when something has changed, either there is a problem with httpd.conf or your virtual hosts. One of my friends enountered an Apache failure and it turned out to be one of its modules got deleted somehow. So the best course of action is to make a copy of your httpd.conf and virtual hosts before you ever change them. If you mess them up you can always replace the old copy to get Apache back up and running.

  • To try and troubleshoot Apache failures, you should take a look in syslog (tail -n 100 /var/log/messages). See if there are any messages from the httpd process, and if so, does it give you an idea of what may be wrong? I’ve been able to track down errors by doing this one thing. Get the error and look for it on Google. It’s unlikely you’re the first person to experience it.
  • If it’s a config file error, you can test your httpd.conf and virtual hosts with httpd -S (or on older Apache versions use httpd -t -D DUMP_VHOSTS). If there is a problem with your files, Apache will tell you when you run the test.
  • Another useful hint is to try starting Apache in debug mode using httpd -e debug and see what it prints to the screen. If there is an error it should be shown to you. Once again, if you see the error, use Google to find an answer.
  • If all else fails, there is trial and error. Move your httpd.conf and virtual host files to another directory and reinstall the Apache RPM’s. Once Apache is back up with default settings, start making changes to the new httpd.conf and virtual hosts from the old files. Each time you make a change, save your changes and restart Apache. Like I said, it’s trial and error and you may be able to track down the problem.

Can I use Active Server Pages (ASP) or ColdFusion (CFML) with Apache on Linux?
Why would you want to? Those are weak and useless languages (my opinion of course). You should use PHP instead, however, if you have the need to use them, there are ports from third party companies that let you use those languages with Apache on Linux. For Active Server Pages, Sun Microsystems has a product called Sun ONE Active Server Pages (a former ChiliSoft product) that let’s you use ASP on Linux. They have a free trial version you can download and try out.ColdFusion is no longer a Macromedia product and is now maintained by Adobe Systems. You can download a free trial version after you register for an account.

[more to be added]

Conclusion

I think I may have went overboard with all the information in this Apache tutorial because it took me four days off and on to prepare this document. I tried to be as thorough as possible but I realize I may have missed some of the little things like in the Q&A section above. Nevertheless, you now have the skills to configure Apache from the ground up from editing httpd.conf, creating a structured web site directory schema, and adding HTTP and SSL virtual hosts. You also learned a little bit about Secure Sockets Layer and SSL web server certificates. In case you didn’t know, if you run into problems or ever need any help you can always post your questions to the Linux category of the forum and I will try and respond to them as as possible.

 
2 phản hồi

Posted by trên 10/12/2011 in Fedora + CentOs

 

Install Centos 5 Server Setup CD1 by graphic mode


các bạn xem video này nhé!

 

 

nguồn: youtube

 

install Centos 5 Server Setup CD1 and text mode


In this tutorial i will show you how to install Centos 5 using the first CD only as base system for your Linux Server.
Select one mirror from http://mirror.centos.org/centos/5/isos/ and download CentOS-5.2-i386-bin-1of6.iso then burn the iso file in your favorite program, this will produce Centos bootable CD
change your BIOS setting to make boot from CD

if link not work pick other mirror from here

http://www.centos.org/modules/tinycontent/index.php?id=30

this will work with centos 5.3 and centos 5.4

let’s watch the tutorial now

Install Centos 5 Server Setup CD1 and text mode

sau khi clink vào link trên,nếu ko thấy gì,các bạn vui lòng f5 lại trang vừa mới mở!

Nguồn: ST

 

Configure Windows And Linux Central Authentication With LDAP (Not Secure)


The Lightweight Directory Access Protocol (LDAP) is an application protocol for accessing and maintaining distributed directory information services over an Internet Protocol (IP) network. This service allow you to store your users and passwords in a central location. Now, I will guide you how to setup network authentication with LDAP.

1- Prepare LDAP Server
Step 1. Install required LDAP Server packages
# yum install -y openldap openldap-servers openldap-clients openldap-devel nss_ldap
Step 2. Create directory to store LDAP database. This directory will hold all the bdb database files.
# mkdir /var/lib/ldap/itbox4vn.com
# chown ldap:ldap /var/lib/ldap/itbox4vn.com
Step 3. Create ldap admin password
Step 4. Edit /etc/openldap/slapd.conf
database         bdb
suffix           “dc=itbox4vn,dc=com”
rootdn           “cn=Manager,dc=itbox4vn,dc=com”
rootpw        {SSHA}rwfbf8wrOQiXtbzWQeYBWu97RPju+7kH
directory        /var/lib/ldap/itbox4vn.com
Step 5. Create a test account named ldapuser
Step 6. Create LDAP Database
The password conversion script is named migrate_common.ph. Edit the file and replace all instances of the string “padl” with the string “itbox4vn”.
# vi /usr/share/openldap/migration/migrate_common.ph
# updatedb
# cp /etc/openldap/DB_CONFIG.example /var/lib/ldap/itbox4vn.com/DB_CONFIG
# /usr/share/openldap/migration/migrate_all_offline.sh
Step 7. Start the LDAP Server
2- Configure Linux Clients
Step 1. Install the necessary packages:
# yum install -y openldap nss_ldap openldap_clients openldap-devel
Step 2. Choose LDAP authentication
# authconfig-tui
The /etc/nsswitch.conf and /etc/ldap.conf file is automatically edited for you. There is no LDAP client service that needs to be started. When the config file is in place, you are all set to start authenticating via the network. When troubleshooting network client authentication, check the/var/log/secure file for the errors.
Step 3. Create The Home Directory For ldapuser On The LDAP Client
Step 4. Login and change password of ldapuser
3- Configure Windows Clients
pGina and LDAPAuth plugin are very useful applications for Windows System join LDAP Server. You can down load these from http://www.pgina.org.
After installation, we start to configure:
Now we restart Windows System and login to LDAP Server. I created a user called ldapwin for Windows using.
4- Common LDAP Administrative Tasks
Step 1. Change user password
Step 2. Create Modify LDAP User Script
#!/bin/bash
grep $1 /etc/passwd > /tmp/modifyldapuser.tmp
/usr/share/openldap/migration/migrate_passwd.pl /tmp/modifyldapuser.tmp /tmp/modifyldapuser.ldif.tmp
cat /tmp/modifyldapuser.ldif.tmp > /tmp/modifyldapuser.ldif
ldapmodify -x -D “cn=Manager,dc=itbox4vn,dc=com” -W -f /tmp/modifyldapuser.ldif
rm -f /tmp/modifyldapuser.*
Step 3. Create Add User Script
#!/bin/bash
grep $1 /etc/passwd > /tmp/changeldappasswd.tmp
/usr/share/openldap/migration/migrate_passwd.pl /tmp/changeldappasswd.tmp /tmp/changeldappasswd.ldif.tmp
cat /tmp/changeldappasswd.ldif.tmp > /tmp/changeldappasswd.ldif
ldapadd -x -D “cn=Manager,dc=itbox4vn,dc=com” -W -f /tmp/changeldappasswd.ldif
rm -f /tmp/changeldappasswd.*
Step 4. Create Delete User Script
#!/bin/bash
ldapdelete -x -W -D “cn=Manager,dc=itbox4vn,dc=com” “uid=$1,ou=People,dc=itbox4vn,dc=com”
Have fun!
 

Monitor System Performance In Linux


Every time a program or command is run, a process is created for it. These processes are all unique and identified by the process identification (PID) that becomes allocated to it. Management of processes can help keep the system stable or help when the system becomes unstable. Here are some of the process management commands you use:

ps           Displays information about running processes
kill         Terminates a process
pgrep Finds a process based on its PID
pidof Displays all processes related to service or command
top         Monitors system resources
renice  Adjust the priority of a particular process
1. View Root Processes
To view process with more detailed information of the processes the root user currently owns, you can use the “ps u” command. Here, you see not only the PID, but also the CPU and memory utilization. A common set of options that usually use when working with the ps command is aux combined with the grepcommand to get detailed information about a particular process.
2. Terminate The Process
To terminate a process, we need to know the PID of the process. For example, if you want to stop the SSH service because it isn’t responding, you just have to look for the PID associated with the SSH daemon. As you can see, the PID of SSH service is 2162. To kill the process forcefully and effectively stop the SSH service, you can do the following:
# kill 2162
There are two more other commands you can use to determine the PID(s) of a service or command.
3. View System Process
With the top command you will have an overview of processes on the system, including memory usage, CPU utilization, and more.
4. Change The Priority Of Process
You can use the renice command to give that particular process higher priority on the CPU. The priority values range from -20 (first priority) to 20 (dead last priority). Only the root user may set processes to use a priority under 0. For example, I change the priority of SSH service from 0 (default) to -2.
Have fun!
 

Basic Linux System Security


Understanding the basics of Linux security is the best method to keep your business from outside or inside attacks.

Security Through TCP Wrappers
TCP Wrappers is a host service that can be used to limit or control access from remote host. First, check for the TCP Wrappers package to make sure it is installed:
You can limit access to either users or hosts via the /etc/hosts.allow and/etc/hosts.deny files. The TCP Wrappers service scans the two files in the following:
+ Search the hosts.allow file.
+ Search the hosts.deny file.
+ If not found in either, allow.
For security, you should deny all access to the SSH service (/etc/hosts.deny):
sshd : all
Suppose that you want to restrict access to the SSH service and allow connection only from the local network except host 192.168.1.40 . In this case, you would use the following rule in the /etc/hosts.allow:
sshd : 192.168.1. EXCEPT 192.168.1.40
If you want to allow only a single IP address to be able to access the server, you could do something like this:
sshd : 192.168.1.2
You can restrict to hosts from particular domains with the following:
sshd : .itbox4vn.com
With this line in your /etc/hosts.allow file, you allow any system on theitbox4vn.com domain to access this server. When troubleshooting TCP Wrappers, you can use the /var/log/secure file to view any information that is recorded.
Firewall Rules Using iptables
Managing the firewall is essential because many services depend on being able to interact with the outside world or the rest of your network’s security. Because the firewall is set up be default, you just need to verify that package is set to start when the system boots up:
During the boot, the /etc/rc.d/init.d/iptables script executes and starts service with the rules found in /etc/sysconfig/iptables.
Before you start configuring rules, you can view any existing firewall rules:
Let’s look at a basic iptables example to see how a rule is created. Allow SSH connections over TCP port 22:
# iptables -I INPUT -p tcp -m tcp –dport 22 -j ACCEPT
You can see that it is inserting this rule (I); using the default input chain (INPUT); matching only TCP connections (-m tcp); using the TCP protocol (-p tcp); looking for incoming connections on port 22 (–port 22); and, if a packet is found, jumping (-j) to the acceptant chain (ACCEPT) to allow the packet.
When Troubleshooting Firewall Rules and connections, you can temporarily disable the firewall to make sure that the firewall is really what is causing your problem in the first place. You can use the status option of the service command to view the status of iptables.
# service iptables status
Have fun!
 

Working With Security Enhanced Linux (SELinux)


Security Enhanced Linux (SELinux) is another layer of security for the Linux OS. Instead of turning it off, however, you could use SELinux in “permissive” mode, which allows everything to function normally but logs warnings when actions or commands would have been blocked.

1. Understanding SELinux
Let’s start with some of the basics to understanding SELinux. It can run in three different modes:
disabled          SELinux is turned off and doesn’t restrict anything
permissive SELinux is turned on, logs warnings only
enforcing SELinux is turned on and blocks actions related to services
View the current status of SELinux
2. Configuring SELinux
You can change the mode in which SELinux operates by using setupcommand or changing the config file (/etc/selinux/config)
3. File Contexts
SELinux uses three different contexts to enforce security: userrole anddomain (also called type).
User:
unconfined_u Unprotected user
system_u         System user
user_u Normal user
Role:
object_r File
system_r Users and processes
Domain:
unconfined_t Unprotected file or process
Let’s take a look at SSH service as an example.
The first field you see here is system_u, which is a system user. The second file contains system_r, which is a process. The third field showsunconfined_t as the domain.
Another example is ssh_config file:
You see the user is system_u (a system user), the role is object_r (a file), and the domain is etc_t. Any service that has access to the ect_t domain is able to access this file.
To change the context of a file or directory, you can use the chcon command:
Step 1. Change the user context from normal user to system user:
Step 2. Reset the context of your file back to its original context:
4. Service and Boolean Options
Each service controls certain actions with a set of options defined as Boolean values (on or off). This session will help you how to enable access to different services.
To view these Boolean options, you can use the getsebool command combined with grep to look for specific options.
After deciding which Boolean you’d like to change, you need to enable or disable the value appropriately. To achieve this, you can use the setseboolcommand. Suppose you want to be able to access your home directory through the web server. You need to adjust the httpd_enable_homedirsBoolean.
5. SELinux Troubleshooting
There are few packages that you should install first:
# yum install -y selinux-policy setroubleshoot-server
You will also find the following two log files handy:
/var/log/audit/audit.log  Logs SELinux denials
/var/log/messages Logs SELinux denials
Two common commands you can use to hunt for error messages include:
# grep “SELinux is preventing” /var/log/messages 
# grep “denied” /var/log/audit/audit.log                  
Problems can arise in SELinux for numberous reasons. However, the top three include:
Labeling problems: Using a nonstandard directory tends to cause problems if the directory or files aren’t labeled correctly.
Correct context: When you’re moving files, they can lose or retain incorrect contexts, causing access errors. Use the matchpathcon command to verify the correct context.
Confined service: If certain Booleans are not enable, a service may have trouble operating or communicating with other services.
Have fun!
 

Backup and Restore Linux OS


Backup your Linux OS is the first thing I want to do after installing my Linux machine and necessary softwares. You can install your OS again but it takes long time and effort to install all the stuffs in your machine. Now I start to backup and restore my Ubuntu 9.10. It’s the same to the other distributions.

Before we start, let’s make sure that you have root permission to do all the things.
Step 1. Start to backup:
# tar cvpzf backup.tgz –exclude=/proc –exclude=/lost+found –exclude=/backup.tgz –exclude=/mnt –exclude=/sys /
The command above creates a compress file of all the system except from/proc, /lost+found/mnt and /sys directories. However, you need to unmount all partitions such as flash drive, network drive…, if you don’t want to run out of memory for your backup file.
It takes about 5-10’ to backup your system. For better compression, you can use the command below to backup your OS (take more time).
# tar cvpjf backup.tar.bz2 –exclude=/proc –exclude=/lost+found –exclude=/backup.tar.bz2 –exclude=/mnt –exclude=/sys /
  
Step 2. Restore your machine:
If you want to restore your OS just type the following command in the command line.
# tar xvpfz backup.tgz -C /
OR
# tar xvpfj backup.tar.bz2 -C /
In worst cases, you can’t get into your machine. Let’s use live cd (Ubuntu, CentOS…) to get into the command line. Then decompress your backup file to restore your system.
* Note: There are some other ways that may work when backup your system. First, I think you can use Norton Ghost (Hirent’s Boot Disk) to ghost your whole disk. Second, you can copy your hard disk to another disk (http://www.itbox4vn.com/2011/06/how-to-copy-content-of-one-hard-disk-to.html).
Have fun!
 

How To Secure SSH Connection


Secure your SSH connection is very important. I guess that you don’t want your password being hacked, right? There are some tips that can help you to have a securer connection. Here we go!

Step 1. SELinux restrictions for SSH

Step 2. Use TCP Wrappers to limit the hosts that can connect to server:

# echo “sshd: 192.168.1.” >> /etc/hosts.allow
# echo “ALL: ALL” >> /etc/hosts.deny            

This allow all clients with the 192.168.1.0/24 subnet to connect into the SSH server, and it disallows any other host outside this subnet.

Step 3. Change the options to improve security (/etc/ssh/sshd-config):

When you change these options, the default port isn’t known to everyone, and only the internal network adapter listens for connections. Then only the two users (root and user01) are allowed to connect to the SSH server.

Have fun!

 

Setting Up VNC Server On Linux (RHEL/CentOS)


The VNC server allows you to remote into the user’s system and view her desktop. With the end user’s desktop in view, you can more easily troubleshoot any issues she is having. Setting up a VNC serer isn’t hard at all.

1. Setting VNC Server
Step 1. Install the vnc-server package:
# yum install -y vnc-server  (RHEL6 is tigervnc-server)
Step 2. View the default line items:
The first line defines the user who is allowed to log in to the system. The second line lists arguments passed to the VNC server when the service starts.
-geometry Defines the site of the viewer when the client connects
-nolisten tcp Denies TCP connections to the VNC server
-nohttpd         Denies web VNC clients from connecting
-localhost Forces the use of a secure gateway (port forwarding)
The number that appears (2 in this case) is the number of the session for defined user. The VNC server runs on port 5900, but the actual port that will used is 5900 + the number defined = 5902 in this case.
Step 3. Change your config file to the following:
Step 4. Set the password:
# vncpasswd
Step 5. Start the VNC server:
Note:  If you’re using RHEL5, you need to do some adjustment.~/.vnc/xstartup file to uncomment the first two lines:
Then stop and restart VNC server
2. Setting up VNC clients
Step 1. Install the VNC package that contains the client software:
# yum install -y vnc (RHEL5)
OR
# yum install -y tigervnc (RHEL6)
Step 2. Connect to the VNC server:
The system now prompts you for the password that you set up for the user.
Have fun!