Switching Run Levels in CentOS

I am assuming that anyone reading this is at least familar with run levels and linux but please have an understanding of what they are and how they will affect your system(s).

Runlevels

0 - halt (Do NOT set initdefault to this)

1 - Single user mode

2 - Multiuser, without NFS (The same as 3, if you do not have networking)

3 - Full multiuser mode

4 - unused or Admin

5 - X11

6 - reboot (Do NOT set initdefault to this)

The most commonly used runlevels in CentOS are 0, 1, 3, 5 and 6. Most systems will boot into runlevel 5 with a GUI login and either Gnome or KDE as a window manager running on top of X. This is exactly what someone using the computer as a desktop will want, but for a server you will want to boot to into runlevel 3 (Full multiuser mode). From there you may choose to "startx" manually once logged in. (A common setup for me. I like the ability to have a GUI, but I don't want it running all of the time. Runlevel1 (Single user mode) has been very handy, for instance if you have forgotten your root password or are having trouble booting for any number of reasons.

To change the runlevel of the server upon boot up edit the /etc/inittab

sudo vi /etc/inittab

Around line 18 you will see a line as shown below.

id:5:initdefault:

You simply change the "5" in this case to the runlevel you desire. (In my case 3) Save the file and exit.

Spiceworks releases 3.5!

Just found out that Spiceworks has been updated to 3.5. I never got the chance I wanted to beta test 3.5, but now I don't have to worry about it. Looking forward to the Network Bandwidth Analyzer and the Nagios integration.

I plan on doing a backup of my database tonight so I can do the update after hours. Wish me luck.

Configuring PuTTY to use PKI (passwordless) authentication

Since this is about using PuTTY I would recommend you download it if you haven't already. Also you will need to get PuTTYgen, it might be worth it to grab pscp as well. I like to keep putty.exe and puttygen.exe in my Windows directory, so I don't have to update my path, thus allowing me to run putty right from the command prompt.

Now run PuTTYGen and create a new pair of keys by clicking the “Generate” button. You will have to move your mouse around in the box to generate randomness, so keep doing that until the progress bar fills up. You can keep all the options at their default settings. Then, save both public and private key to a safe location. Name your public key [your_key_name].pub and the private key [your_key_name].ppk.

Now, upload your public key to a directory on your remote system. I used pscp to do this quickly to all of my servers (pscp [your_key_name.pub] user@remotesystem:)

Now you have to import your public key into the authorized_keys file (and authorized_keys2)

ssh-keygen -i -f [your_key_name.pub] >> .ssh/authorized_keys && ssh-keygen -i -f [your_key_name.pub] >> .ssh/authorized_keys2

Replace [your_key_name.pub] with the path to your key. Now log out and start PuTTY.

In Putty, you have to configure the following items:

In Connection/Data, add your remote user name

In Connection/SSH/Auth, browse to your private key file (.ppk)

In Sessions, fill in the FQDN or IP address of your remote machine, give your session a name [session_name] and click on Save.

Now you can use putty to SSH into your remote boxes with out a password. If you are a fan of having one or two click shortcuts create a shortcut to %windir%\putty.exe -load [session_name]. If you gave your [session_name] a name with spaces use double quotes to encase it like putty.exe -load "session name".

Configuring CentOS 5.2 to accept passwordless authentication via PKI

I am sure most people reading this blog know how to do this, setup password less authentication using Public Key Infrastructure, but there are a few minutea that I was missing.

The .ssh directory needs to be read/writeable/executabe by the owner only (chmod 700 .ssh)

authorized_keys and authorized_keys2 need to be read/writeable by the owner only (chmod 600 authorized_key*).

On CentOS 5.2 I also dropped in a .config file into the .ssh directory I was able to connect with Putty and not use a password.

nagstamon: A Nagios system tray monitor

First a little bit about my setup. While its true that I love nagios I don't always want to wait for e-mails to come before I know about a problem. I run a dual PC setup with Synergy2 to connect the two monitors and use only one keyboard and mouse. I keep the second monitor on my helpdesk tickets and nagios service detail page. I also use the Nagios Checker for Firefox on my second screen, but since I run Chrome as my main browser I cannot use that option on my main screen.

Today I found nagstamon and instantly fell in love with it! Configuration is easy. Below you can see what settings I am using for optimal performance in my opinion. Of course nagios-server is my actual FQDN for my nagios server and I don't use nagiosadmin to login to my nagios server.

If you notice on the last tab "Executables" I keep putty in the Windows directory so I don't have to update my path. This allows me to run putty right from the command line anywhere I am once I am at a command prompt.

Also if you don't choose to put NagStaMon in the system tray it will float in it's own tiny window like you see below. Look near the upper left corner by the Firefox Icon.

How-To: Install, Configure, and Import your first SVN (Subversion) repo on CentOS 5

This might only be useful to me, I don't know but at least I will know its out there and when I have to do this again I will know what to do.

Use wget to pull down the latest source.

tar xzfv subversion.1.2.3.4.tar.gz

cd subversion.1.2.3.4/

./configure

make && make install

svnadmin create --fs-type fsfs /path/to/repo

touch /path/to/repo/.htpasswd

htpasswd /path/to/repo username

svn import -m "First Import" --username=username /source/path/ file:///path/to/repo

Setup the apache config file. Since I use named virtual hosts and typically have multple repos here is a sample.

<VirtualHost *:80>

        ServerName svn.yourhost.com

        DocumentRoot "/var/www/html"

        ServerPath /html/

        DirectoryIndex index.php index.htm

        <Location /svn/repo1>

                DAV svn

                SVNPath /path/to/repo/repo1

                AuthType Basic

                AuthName "repo1 Repository"

                AuthzSVNAccessFile /path/to/svn-acl-conf

                AuthUserFile /path/to/repo/repo1/.htpasswd

                Require valid-user

        </Location>

        <Location /svn/repo2>

                DAV svn

                SVNPath /path/to/repo/repo2

                AuthType Basic

                AuthName "repo2 Repository"

                AuthzSVNAccessFile /path/to/svn-acl-conf

                AuthUserFile /path/to/repo/repo1/.htpasswd

                Require valid-user

        </Location>

        <Location /svn/repo3>

                DAV svn

                SVNPath /path/to/repo/repo3

                AuthType Basic

                AuthName "repo3 Repository"

                AuthzSVNAccessFile /path/to/svn-acl-conf

                AuthUserFile /path/to/repo/repo3/.htpasswd

                Require valid-user

        </Location>

</VirtualHost>

The /path/to/svn-acl-conf file contains

[repo1:/]

username = rw

[repo2:/]

username = rw

[repo3:/]

username =rw

This file needs to be updated everytime a new repository is added with the appropriate username and permissions.

I am sure I have missed some steps with this, but it's a decent starting point.

Adding a Virtual Host to Apache2 based on Name in CentOS 5

Well today I was setting up our new web development server and I wanted to use Host / Name based virtual hosts in apache, instead of the previous use of ports.

Much to my amazment this one is really simple.

Make sure to uncomment or add in:

NameVirtualHost *:80

Since I don't want to add multple IP's to the server and I always want to use port 80 (The default for http). 

Then I added 

<VirtualHost *:80>

        ServerName scorpio.mydomain.local

        ServerAlias scorpio

        DocumentRoot "/var/www/html"

        ServerPath /html/

        DirectoryIndex index.php index.htm

</VirtualHost>

<VirtualHost *:80>

        ServerName dev.mydomain.local

ServerAlias www.mydowmain.local mydomain.local

        DocumentRoot "/var/www/html"

        ServerPath /html/

        DirectoryIndex index.php index.htm

</VirtualHost>

<VirtualHost *:80>

        ServerName dev.myotherdomain.local

ServerAlias www.myotherdowmain.local myotherdomain.local

        DocumentRoot "/var/www/html"

        ServerPath /html/

        DirectoryIndex index.php index.htm

</VirtualHost>

<VirtualHost *:80>

        ServerName dev.mythirddomain.local

ServerAlias www.mythirddowmain.local mythirddomain.local

        DocumentRoot "/var/www/html"

        ServerPath /html/

        DirectoryIndex index.php index.htm

</VirtualHost>

When it was all said and done this was pretty easy.  Not as easy as doing the same thing in IIS in my opinion but still very easy.  I could have tried this with Webmin but I did this all using vi from the CLI.