I had for a long time a serious problem with my SSH server on one of my machines. It used to respond to the connection requests very slowly, usually after passing 20-30 seconds. This was very annoying, since it delayed my workflow, and also, because only one of my machines that run the SSH server had this behavior.
After searching around the Internet, I was lucky this week and found the correct answer on this problem. Many suggested various tricks and customizations, but the problem was very simple indeed.
It had to do with the so-called “Reverse DNS Lookup” (or something similar). In few words, sshd tries to look up the remote host name and check that the resolved host name maps back to the very same IP address. It is a security feature, yet slows down the operation of the server.
This is a default action for newer versions of sshd (and that’s the reason older machines haven’t this behavior) and the problem is that it isn’t mentioned in the appropriate config file. So, adding the line
UseDNS no
in the /etc/ssh/sshd_config file solves the very annoying problem.
So simple!
I found the way to control the size of the ~/.thumbnails directory at Gnome.
The ~/.thumbnails directory is used in Gnome by Nautilus to store small pictures (thumbnails) of various files. The problem is that it can become quite large especially when limited space exists on /home directory.
You can control the behaviour of the thumbnails with the values of /desktop/gnome/thumbnail_cache directory at Gconf. For example, I changed to a smaller cache size by issuing the command:
gconftool -st int /desktop/gnome/thumbnail_cache/maximum_size 128
In this article, I found a great tutorial of how one can set up a working environment between two or more people at different geographical areas. The solution is to use screen and ssh. Read more…
This post contains content and examples originally found at an article on ServerWatch. It contains some notes and tips about the Debian packaging system, with some commands beyond the usual, everyday ones that somebody might use. Read more…
apt is a debian beast for me, with a lot of power and capabilities. Unfortunately, I feel that I don’t really and fully understand its capabilities. After skimming at its manual, I found the following very useful information.
apt-get cleanwill remove ALL the cache of packages apt has downloaded. These packages are cached in the system to be re-used in case of re-install or copy to another machine.apt-get autocleanremoves only package files that can no longer be downloaded, i.e., old package versions or obsolete packages.apt-get update -uwill do the same job as Gnome’s Update Manager, only run at the command line.
I’ve read recently the following article on ServerWatch, where I found some advanced techniques for handling RPMs. Although I have no access to any RPM-based machine at the moment, nevertheless it is always useful to keep handy this information for future reference. Read more…
This note is something I always forget when it comes up and I always find it difficult and time-consuming to recover from the manual and documentation. Let’s hope the next time I’ll need it, it would be much easier for me to spot it!
It has to do with a feature of GNU Automake, one of the tools used in the GNU Building System, a toolchain for building software under the normal Open-Source practices. This feature is called DESTDIR and it allows somebody to checkout a software installation without actually installing the software.
Here is an example of the usual build procedure I follow:
cd src # enter program's source directory pushd /tmp/build # go into a temporary build directory ~1/configure --help # first, check available config options ~1/configure make make DESTDIR=/tmp/inst install make install
Now, you can find in the /tmp/inst directory all the files from the package that are going to be installed, preserving even the directory hierarchy of the installation.
Hello, my beloved Blog!
Today, I came to the world of blogging and WordPress as thousands have done in the past. The content of this blog will be mostly some technical notes on Computer Programming and Software Development, Linux, Gnome, System Administration.
My aim is not to publish any new ideas or revolutionary algorithms. Instead, I would like to keep some notes about the problems I find in my way and the solutions I found about them for future reference. After all, most of the time we face the same problems again and again, and sometimes we can’t recall the method we used to overcome this obstacle. So, most of the staff in this blog will be based on other blogs, books or articles.
Have a nice journey!
Welcome to WordPress.com.
I have to write some things about the blog’s title and explain its name. The “break” at the blog’s title doesn’t mean I want to crack Linux. Instead it comes from the familiar to many command given as sample below:
(gdb) break main
Unfortunately, we can’t speak for a “main” function in Linux. Instead, I need to have a “breakpoint” at Linux in order to “debug” it and understand many issues hidden behind the hoods.
Let the journey begin!
break linux 