Linux Blog

@nongeekboy on Twitter tweeted? about a blog post: Simple Script To List Groups in passwd File. I read it and have done something similar before so I figured I’d blog it here so I won’t have to write it again. Anyway, along with the point of this post, since this is suppose to be quick.

A question that is often asked is “What groups am I in?”

The easiest way to find out is to type the command: groups
This will give you a list of the groups you are in separated by a space. There are some other fancy ways of getting the groups but they rely on the `id` command. Running groups with no user name, its the same as running id -Gn.

Here are some other variations that you can try if you need to script the output:

groups
id
id -G
id -g
id -n
id -nG <user>
and the obvious:
id –help
info id

Bookmark It

Run levels in Linux are a great thing. Basically, a run level is by definition a configuration for a group of processes. The run levels and default run level is specified in /etc/inittab. Most Linux systems these days, with exception of a few boot into run level 5 which is generally a graphical user interface such as KDM or GDM. The others boot into run level 3 most servers will boot into this run level which is multi-user with networking but no X, and is many users preference.

To define what run level your system boots into by default you would edit the /etc/inittab file and edit the line similar to:

id:5:initdefault:

This is run level 5, if you wanted to switch to command line you’d change the 5 to 3 and vice versa.

If your not ready to make the jump yet but would like to check it out, you can (as root) use the command telinit to tell init to change run level. If you are in run level 5, try (be prepared to lose everything in X, as it will kill everything for you)

telinit 3

If you are doing maintenance, you may want to switch to level 1 which is single user mode. Level 2 on Fedora is the same as 3 except it doesn’t have NFS support.

Level 0 is halt and run level 6 is reboot which are the best ones to accidentally set as a default run level (trust me on this one.) For more information on the different run levels check out the man pages.

Bookmark It

I think its pretty evident that I love sudo right? Well, what I do not like about sudo is the timeout. I understand while its there but five minutes is not exactly what I’d call an overly generous time, especially when I’m parked here at my desk for hours upon end. This tutorial shows you the line you’ll need adjust the sudo timeout:

First as root you’ll want to get into the sudo file and edit it. I’m sure you know how to do this since you’ve probably already visudo’ed your way into using sudo and are now trying to adjust the timeout. For those just reading for the sake of it, you’ll do the command: visudo

Right, now you’re there, you’ll either be in nano, pico or vi depending on your distribution. Search for the Defaults section, and put

Defaults:<your username> timestamp_timeout=<your timeout>

Replace your username with yours. Change your timeout to the number of minutes, or -1 for unlimited per session. Save and quit, then exit. Try it again, then try it again after the sudo timeout you set has changed. If it works, great news if not double check your sudoers file for another Default property that may be acting up.

Bookmark It

Here is another quick tutorial;

Some times its nice to tunnel through SSH. Perhaps you have SSH running but the firewall does not allow anything but SSH in. You can tunnel VNC (or any other service) through SSH by doing the following:

On the machine local to you establish an SSH connection to the remote machine with “Local (-L)”  port forwarding. This may seem confusing and often confuses me, where <-p PORT> is optional

 ssh -L 5901:localhost:5900 username@HOST <-p PORT>

Once I have the connection established I can now use vncviewer to connect to my local host with the port specified

vncviewer  localhost:5901

Thats all there is to it, have fun!

Bookmark It

So, I hear you need to use Tomcat on Fedora eh? Not happy with the available Tomcat version from the repository? Well my friends you can add a custom Tomcat to Fedora and have it run as a service.

This post is somewhat related to: my Adding a service on Fedora post except this one is more specific to Tomcat. If you’d like more information on adding services to Fedora that is the place to look.

Here is the script that I have been using: (Read on …)

Bookmark It

There is no doubt about it that SSH is a handy administration and remote access tool. Have you ever wanted to add port forwarding or cancel port forwarding from within an active SSH connection? Well, a feature that many do not know about is the ssh escape feature. With this pseudo terminal you can do some nifty things, such as forward ports over an active SSH connection.

To do this, follow these instructions:
(Read on …)

Bookmark It

Linux users often like to boast about their awesome bootup times. I thought that there was nothing cooler than getting a wicked fast bootup time, until now. A while back I found this nifty application called bootchart and shoved it in my bookmarks. I was randomly surfing my bookmarks, came across it again and gave it another shot. (Read on …)

Bookmark It

Lets say you use subversion on your home PC to keep track of projects and you want to checkout or export your project from a remote location. Here’s the catch, sshd is running on a custom port or forwarded from another. For some reason the command line SVN client does not support a port parameter when using the defacto svn+ssh://

svn co svn+ssh://thelinuxblog.com/owen/svn/project1/trunk project1
ssh: connect to host thelinuxblog.com port 22: Connection refused

Well we know why the error above happens its because I happen to run SSH on port 1337. The following work around requires root privileges, and may mess with your system a bit but if you really need to check something out, then it will work.

As root, login and stop SSH if you run it. With SSH Forward port 22 with a local SSH forwarding connection to the remote host.

[owen@thelinuxblog.com]$ sudo su -
[root@thelinuxblog.com]$ /sbin/service sshd stop
[root@thelinuxblog.com]$ ssh -p 1337 owen@thelinuxblog.com-L 22:<internal ip>:1337

Once this is done, your localhost:22 now forwards to your remote host. With another session (on your local machine) you can verify the connection by using ssh localhost. You will probably get warnings about the hosts identity being changed, or not verifying it, but you can ignore then. Once you’ve tested it, just use SVN as normal. When finished, remember to logout of the SSH session, and start SSH back up again if you run it.

Bookmark It

TheLinuxBlog.com was intended to be a blog where I could log my thoughts, findings and generally keep track of how to do stuff. Since it was started this is still the goal. Ultimately I’d like to refer to TheLinuxBlog on how to do something just as much as I refer to google for everything else. That being said, one thing that recently came up for me was “How do I change my desktop on Fedora”.

I had done this before but I couldn’t remember the command to do so. Well, since this blog is as much for me as it is for the reader I figure I can post the how to here and kill two birds out with one blog post, I mean stone.

The program I use to change my window manager on Fedora is: switchdesk.

Switchdesk can be installed by Yum or if you installed Fedora from DVD or CD and didn’t fine tune your packages then you probably have it already. All you have to do to run it is type:

switchdesk

Now, if you are in an X session you will get a nice graphical dialog that will help you change your desktop manager. If your at the terminal it will exit and ask you nicely to type either gnome, kde, xfce or any other window manager you may have installed.

Don’t ask me why every distribution has a different named command and interface to achieve the same thing thats just the way it is. Maybe one day I’ll get a list of all of the commands and post them. Alternatively if anyone wants to start a list feel free to post them in comments or by e-mail.

Bookmark It

This is a common problem with the Apache web server. It is considered a security risk to show the directory contents by default. Especially those in the root directory. The way to fix this is pretty simple, you first need to make sure that there are no -Index directives in any of the configuration files. In your httpd configuration directory grep with a line number for Options like so:

grep -n Options *

If you don’t see any -Indexes then its possible there are included files. Check these directories if you know them, if you don’t grep your configuration file for the Includes:

grep -n Include *

If you’re still getting the test page or a permission denied error after removing the directive that disables Indexes then it may be a permissions issue. Apache needs +x access for all users to enable directory listings. Change the permissions and it should work.

On Fedora there is a file called welcome.conf in the conf.d directory. This can be removed and if you have Options All set, then you should be good to go. On other distributions like Slackware it is easier to accomplish. I wouldn’t recommend allowing any sort of directory listing in a production environment but in my development case, where I only allow access to my test server on a per host basis this is not a big problem.

Also, remember to restart the server after making changes.

Hope this helps, if it doesn’t be sure to drop a comment!

Bookmark It

Have you ever wanted to make a command for something that did not exist? Perhaps modify the functionality of a particular application to something more suitable? I know I have. For example, and I know that there is probably a better way to do this, but on certain Linux machines (such as servers), I like to clear the output before exiting. While I can type the command:

clear; exit

this still leaves me with a line at the top of the screen. Some distributions clear this automatically, for those that don’t an alias can be used.The basic principal for a bash alias is easy. You set an alias up and then use that alias instead of the command.
It appears that aliases have precedence over any already existing applications in the path so it becomes handy if you wish to override a command or perform a task before launching a built in command. I’m sure that this option can be changed if needed.

Now you know what aliases are here is how to use an alias to override the exit command in bash.

alias exit=”clear; exit > /dev/null 2> /dev/null”

Aliases with parameters can get tricky, the best workaround I have found is to write a shell script and put it in your local bin directory.

Bookmark It