Linux.com

Home Linux Community Community Blogs

Community Blogs



Squid and Basic Authentication

This is perhaps the easiest authentication helper to configure in Squid, but also the most insecure. The biggest problem with Basic is it transmits username and password in clear text, hence very susceptible to network sniffing or man in the middle type attacks. The only reason I’m writing about it is it’s a valid authentication mechanism in some limited circumstances. Secondly I want to show you how authentication has evolved over the years. Ultimately you want to Kerberos authentication with your Squid proxy, but before we got there we had basic. And here is how to configure it; First thing that requires out magic touch is Squid’s configuration. Locate and navigate squid.conf The first section you’ll come across is for configuring authentication. It’s called; # OPTIONS FOR AUTHENTICATION # ----------------------------------------------------------------------------- You’ll notice there are many comments in this section explaining all the different options. But let’s jump ahead to what we came here for… Locate the following lines; note they will be commented out. Enable them by removing the hash character ‘#’ auth_param basic program auth_param basic children 5 auth_param basic realm Squid proxy-caching web server auth_param basic credentialsttl 2 hours If you haven’t noticed already the first parameter auth_param basic program configures the location of an external helper program. This helper program is named pam_auth and on an Ubuntu system is located in the /usr/lib/squid directory. In fact all authentication helpers are located in this directory. Therefore our first line should look like this; auth_param basic program /usr/lib/squid/pam_auth Next we have the children parameter. This configures the specified number of processes to handle incoming authentication reuqests. In above example pam_auth will spawn 5 separate processes to handle all authentication requests. Anywhere between 5-10 helper processes is a good starting point. If Squid runs into trouble, it will tell you in /var/log/squid/cache.log , monitor this file closely. Then we have a realm parameter. This is a string which is presented to the user when the authentication prompt appears on screen. With Basic authentication this is an arbitrary string value. You can use anything, like; “Welcome to my really cool Proxy Server. Enter your Username and Password” Lastly we have the credentialsttl parameter which dictates how long Squid caches authentication requests internally. Keep in mind a small value increases Squid load, while a larger value will reduce it. You may need to play with this if you notice your Squid box is really busy. The last piece to this puzzle is enabling Squid’s authentication ACL. This includes changing two additional parameters. ( ACL & HTTP_ACCESS). The default ACL bases access or no access on client subnets. ACL LOCALNET SRC 192.168.0.0/24 is an example of one. To enable authentication, comment out above default ACL and replace with this; acl authenticatedusers proxy_auth REQUIRED Lastly enable above access list, named authenticatedusers http_access allow authenticatedusers That’s it. Restart Squid service and you should now be prompted for user name and password. You session will be authenticated until you close your browser. www.digitalboundary.net/wp
 

KeePass Password Safe - Keep and manage multiple account passwords

I'm sure you've heard numerous times from many sources, web site logins should always be unique. Unfortunately few follow this rule and often reuse passwords among different Internet sites. Primarily because keeping track of unique passwords is a real hassle. Really! think about all the sites you use daily, then double that for occasional sites and before you know it you're managing 50 passwords or more. So What's the solution? You could keep a ledger, but then you'd have to photocopy it a few times for every computer in the house. You could spend money on commercial solutions like 1Password or let the Open Source community help. KeePass Password Safe is a great and easy to use alternative. Best of all, its cross platform and free to use. You can download KeePass from www.keepass.info Installation is your typical Windows clicky-click. Installation has 3 options, Full Install, Compact or Custom. Full install maxes out at 5.4MB, compact just 2.8MB and custom anywhere in between. Full Install includes additional libraries, XML stylesheets and a number of optimizations, compact simply includes core KeePass libraries. After installation, the first required step is the creation of a KeePass database. Your encrypted database will store all recorded passwords and any other confidential information you'd like to keep safe. You have the option of encrypting your database with a master password, Key file, Windows user account or a combination of the three. KeePass uses AES/Rijndael 256bit encryption, so you can be confident encryption is strong. To improve database functionality and performance, tweaks are available in database settings. For example to reduce the chance of password type guessing attacks, AES/Rijndael encryption uses Key transformations. Default setting is 6000 times, but you can increase this value to whatever you like as long as you realize larger values increase database load time and a slight performance hit. Other options are available from a simple description to database compression to reduce file size. Once the database is created you're ready to add your first secret entry. More on that next week. http://www.digitalboundary.net/wp
 

Corks? Or Screw Tops? Why the Experience Matters

I've noticed a disturbing trend amongst a few of the high quality wineries in my state. They have abandoned the cork to close their high-end wine bottles and turned to screw caps. This is good news to people who struggle with how to get a cork out of a wine bottle. 

Read more... Comment (0)
 

Building LFS( Linux From Scratch)

After struggling for nearly one day, the LFS, linux from scratch, is finally built. I would note something about the building process. 1. about ssh There is some time that I just want to copy the command in LFS book and execute it directly. since copy between host OS and guest OS would be troublesome, SSH would be a convenient way to copy, paste and execute. execute "/etc/rc.d/init.d/sshd start" to start ssh service in LFS, one may also need to configure network interface using ifconfig command. 2. about mount After compiling all packages and changing root directory, I try to execute grub-install, and it tells me that there is no hard disk. And I eventually figured out that "mount -v --bind /dev ${LFS}/dev" is very critical in installing grub boot loader. Without it, grub-install would not find hd0.
 

openSUSE Weekly News 192 is out!

I'm happy to announce the new "openSUSE Weekly News, Issue 192".

 

 

In this Issue:

  • openSUSE Conference 2011
  • Plasma Active Status Report
  • Beta Pizza Party
     

You can download it there:

We hope you enjoy the reading :-)

If you want to help us collecting interesting articles for the openSUSE Weekly News, so you can all your stuff into our new ietherpad: http://os-news.ietherpad.com/2.

Found Bugs? Please place it in our Bugtracker: http://developer.berlios.de/bugs/?group_id=12095

Features, Ideas and Improvements can placed in our Featuretracker: http://developer.berlios.de/feature/?group_id=12095

Older content can be found there.

 

GuitarPro6 on Fedora 64 bit

 

Guitar Pro 6 on Fedora x64

GuitarPro6

 

After I've had a lot of trouble to get this working, I will show you the step by step guidance to run Guitar Pro 6 on a 64 bit version of Fedora (it should be almost the same on  all 64 bit versions of Linux though -> Notice, if you're using a x64 Debian based distribution, it's easier by using getlibs)

 

 

 

1.) First download the guitar pro 6 installer for linux from the website www.guitar-pro.com.

2.) The downloaded file is in .deb container format (for Debian distributions). Simple extract the deb file, and open the extracted folder.

3.) You can find two tar.gz files inside of it: control.tar.gz and data.tar.gz as well as debian-binary. The only file which we need is data.tar.gz, so delete the other two files.

Extract the data.tar.gz file.

Corresponding Shell-Command: tar -xf data.tar.gz

 

4.) Open up the data folder - again you see two folders opt and usr. Open the opt folder and copy the GuitarPro6 folder inside of it to /opt/. (You need sudo rights to do this, so either you do it directly over the terminal, or you call nautilus with root privileges: sudo nautilus /opt)

 

Corresponding Shell-Command: sudo cp -R GuitarPro6 /opt/

 

 

Ok basically that's it, but Guitar Pro 6 doesn't start up yet. It needs some dependencies. So let's see which ones it needs. Open up a terminal and change the directory to /opt/GuitarPro6

Corresponding Shell-Command: cp /opt/GuitarPro6


6.) start the shell script gp-launcher.sh by typing: sh gp-launcher.sh. It should give you the following output:

./GuitarPro: error while loading shared libraries: libportaudio.so.2: cannot open shared object file: No such file or directory

 

So we need all these dependencies (and we need the 32 bit Libraries!!!) So instead of let you find out each dependency after another here are the needed packages you have to install:

 

libstdc++.i686
mesa-libGL.i686
alsa-lib.i686
portaudio.i686
pulseaudio-libs.i686
libXrender.i686
glib2.i686
freetype.i686
fontconfig.i686

libgnomeui.i686
gtk2-engines.i686

 

 

Corresponding Shell-command: sudo yum -y install libstdc++.i686 mesa-libGL.i686 alsa-lib.i686 portaudio.i686 pulseaudio-libs.i686 libXrender.i686 glib2.i686 freetype.i686 fontconfig.i686 libgnomeui.i686 gtk2-engines.i686

 

7.) We're almost done, there's one more problem: After installing these libs and trying to start guitarPro6 with sh gp-launcher.sh the following error will occur:

./GuitarPro: /opt/GuitarPro6/./libz.so.1: version `ZLIB_1.2.3.3' not found (required by /usr/lib/libxml2.so.2)

 

This means GuitarPro6 tries to use a wrong version of libz. The trick is to remove the libz lib in the GuitarPro6 folder and let GuitarPro use the libz version of the system.


Commands:

sudo rm libz.so.1

 

After that GuitarPro6 should be able to start.


8.) Soundbank-Installation:

For some reasons (which I didn't find out yet) the soundbank installation doesn't work. However there's a workaround for this:

 

 

 

  1. Download the file Soundbanks.gpbank directly from the website and copy it to your GuitarPro6 folder (again you need root privileges to do this)

  2. Run sudo /opt/GuitarPro6/GPBankInstaller /opt/GuitarPro6/Soundbanks.gpbank /opt/GuitarPro6/Data/Soundbanks/

     (notice the blanks after each directory argument)

  1. It should sucessfully import it, doing it this way.

 

 

I hope everything worked out fine. You can also add a Desktop link to GuitarPro6 by adding a file in /usr/share/applications.

 

 

If you have questions you can ask here, I'll respond asap: This e-mail address is being protected from spambots. You need JavaScript enabled to view it

 

Buy me a beer :)

 

 

 

 

openSUSE 12.1 milestone 5 (aka, beta 1) released

On September 1st, the openSUSE team released the latest milestone for openSUSE 12.1. This brings the openSUSE 12.1 release closer to replacing standard SysV init with systemd, and closer to GNOME 3.2. Check it out!

 

Newbie help needed! Computing with Linux on USB

G'day Community,

I am a Linux community newbie and this is my first post.  

I'm an engineer and work for a corporation which, despite it's many commendable merits, is constrained by a draconian IT policy that imposes upon us only the option of using an archaic 32-bit Windows XP OS.  Not to mention the wasted productivity and needless cost when a far superior Open Source OS is readily available, being forced to use 32-bit Windows XP is a major impediment preventing me from using the computer resources I have available to run the massive 3D modelling simulation problems I want to in order to better accomplish my work.  I appreciate our local IT guys are only enforcing the policies they've been given, probably without being asked their opinion on them, and I hope to find a work-around solution without rocking the boat.

 

Here's Plan A:

 

I understand Linux can be loaded onto a USB Flash drive and you can run the entire OS from the USB drive without ever actually installing it on the computer's disk.  The programs I want to run only require me to initiate them and direct them to the appropriate data files they'll read with commands through the shell; the programs don't require any GUI or need to be installed.  Once the programs are running, it's just a matter of a massive amount of number crunching for up to a few days using all the available CPU and RAM.  After the process is completed, I could transfer the computed data back to my Windows Laptop on the USB stick and then do all the visualization and analysis with the fancy commercial programs I have for Windows.

 

So, my first question is whether this Plan A is feasible.  I know that Linux would be able to run much better properly installed on the desktop's disk, rather than running it off the USB drive, but how adversely would this affect Linux's ability to fully utilize the computer's RAM and distribute computation among the 4 CPU on the Xeon chip?  It may be worth mentioning that the limiting factor on the size of problems I can run has a lot more to do with the amount of RAM that can be accessed, rather than the time required or speed/number of CPU.

 

The next question is, if running big computational problems with Linux on a USB is feasible and worth pursuing, which would be the best Linux distribution with which to do it?  Does it really matter which 64-bit distribution I'd use, or does it all just depend on the kernel?  I know the merits of different distribution scan bring up controversy and I hope that doesn't distract too much from the main reason for my post.  

Based on the web research I've done so far, I'm leaning towards Linux Mint or Ubuntu because they're both popular and reported to be good choices for newbies.  My friend runs big computational problems like the ones I want to on a Linux Beowulf cluster under CentOS.  He may have gone with CentOS for its network capabilities, which wouldn't apply so much to me because I just want to run the programs on a standalone desktop PC using all of its local CPU and RAM.  Despite that, it makes me think  CentOS is a good way to go just because it's what he's been using to run the same programs for years.

 

I thank the Community in advance for their thoughtful advice and support.

 

Jabood

 

 

 

 

 

 

 

 

Banshee vs Amarok (and the winner is)

If you are like me, you may have already spent a few years running Ubuntu Linux (what ever version) and then to find out the latest version 11.04 with Gnome 3 and Unity just didn't go over that well for ya!

Since I have discovered Ubuntu was not going to be my future operating system, I started to look around and trust me, I have found what I think to be a very good alternative to Ubuntu, it's called Kubuntu.

I am currently working with Kubuntu 11.04 with KDE so far not to many problems, I have had to update a bunch of packages, and go through the normal tweaks anyone would go through with configuring a new operating system to meet their personal needs, but so far have had nothing but good experiences over the last three-weeks.

Until the other day I'm thinking most if not all was looking good, and then I put a music CD in the drive to find out something very unacceptable was about to happened "I couldn't figure out how to play the CD" it took me up to an hour to figure out how to play the music CD I inserted in the drive (I was like WOW) you have to be kidding.

Back in the Ubuntu days with Banshee,  I would insert a CD and had music playing within seconds.  All the tracks would display within a very easy to use media player (it just worked really well) it would also work well with detecting my  iPod Touch (only the photos not the music files yet).

I have researched upgrading my Kubuntu 11.04 with Banshee and so far I see no problems with eliminating Amarok completely, and installing the latest version of Banshee.

In my not so humble opinion, the developers who work on Amarok need to take a few steps back, and rethink what they are doing!

Yes, as you already know, I was used to Banshee with Ubuntu, but when it takes someone that long (an experienced user at that) to figure out how to play a CD-Rom there is something really wrong with the application (the winner with flying colors is "Banshee") trust me, we all listen to music, and for most users I'm sure they want to have an environment that is reliable and easy to use.

 

openSUSE Weekly News 191 is out!

I'm pleased to Announce our new openSUSE Weekly News 191.

 

 

In this Issue:

  • Milestone 5 is out
  • Google Summer of Code
  • Factory update
     

You can download it there:

We hope you enjoy the reading :-)

If you want to help us collecting interesting articles for the openSUSE Weekly News, so you can all your stuff into our new ietherpad: http://os-news.ietherpad.com/2.

Found Bugs? Please place it in our Bugtracker: http://developer.berlios.de/bugs/?group_id=12095

Features, Ideas and Improvements can placed in our Featuretracker: http://developer.berlios.de/feature/?group_id=12095

Older content can be found there.

 

Linux vs Windows Data Servers (the bottom line)

Based on years of writing software within a network environment, and communicating with database servers running on both Microsoft Windows or Linux that provide access to databases like Microsoft SQL server and PostgreSQL  I have to say, Microsoft better look over their shoulders,  the competition is gaining a strong hold on corporate decision makers and how they foresee the future's bottom line.

At this point, the money paid out with setting up a Microsoft SQL data server by the time they factor in the cost of the operating system, the database software, the per-seat licensing, and support you almost have to take a step back and scratch your head with how much money it takes to put this in production.

Again, based on my own experience, I’m not sure why anyone would want their corporate data on a Windows-based server that requires a lot more money and maintenance to keep the server up and running, while maintaining availability, reliability, and security.

A well configured Linux data server provides a much better overall experience for a fraction of the cost.

With how companies are struggling to operate in the green, why anyone wouldn’t explore these open source options like PostgreSQL, MySQL, and operating systems like CentOS just doesn’t make any sense.

 
Page 3 of 122
30 Linux Kernel Developers in 30 Weeks

Who we are ?

The Linux Foundation is a non-profit consortium dedicated to the growth of Linux.

More About the foundation...

Frequent Questions

Join / Linux Training / Board