micrux.net

People who use the Linux/Unix command line to any extent are certainly familiar with the cd (change directory) command. In fact, this is probably one of the first commands you ever learned on the command line. When you’re working with the command line, a lot of times you need to switch to a directory and then come back to the same directory. The cd command allows you to use a dash as an operand for just such a situation.

/var/log $ cd /etc
/etc $ cd -
/var/log
/var/log $

Using cd with a dash instead of a directory location, you can easily jump back to the previous directory. In this example, I started at /var/log and then jumped to /etc and then back to /var/log. It also acts like a toggle, so if you keep doing cd - you will go back and forth between the two most recent directories.

In the Unix/Linux command line environment, exploring the running processes is accomplished chiefly by two programs. top gives you a real-time list of running processes which updates periodically showing the most active processes at the top (hence the name.) ps takes a snapshot of the running processes and dumps them to the command line for you. Both of these methods work well, but a program called htop brings a little bit more sanity and convenience to the situation.

htop is ‘an interactive process-viewer for linux.’ It runs on the command line in real time like top does, but it utilizes ncurses and allows you to browse and interact with the processes. htop also features a colored interface making process management easier and more visually appealing.

htop has easy-to-read resource meters at the top of the interface which make it easy to survey the status of your system at a quick glance. The CPU meter shows you your CPU usage in a way that makes sense to humans. Where top breaks the CPU usage into several different categories, htop combines them into a simple collected usage meter. It is certainly useful to get the very specific CPU usage statistics that top offers, but most people don’t need this much information and are often confused by it. htop does, however, provide some of this information by dividing the meter into different colored sections. Blue for low priority processes, green for normal processes, and red for kernel processes. It also provides the standard CPU load averages, which is another very common representation of CPU load in Unix. htop provides resource meters for memory and swap usage as well. The memory meter is also color coded. Refer to the help section for more specifics about the color codes.

htop allows you to scroll through the list of processes either with the up/down keys, or the page-up/page-down keys. You can specify which attribute you want htop to sort the processes by (e.g. CPU%, Memory Usage, etc.) You can toggle a tree-style display showing you which processes were spawned by which other processes. You can even increase or decrease the nice value of any process. (The nice value is the priority level that the process will receive within the kernel’s scheduler. Simply put, the lower the nice value, the more critical the process and the more CPU time it will be given. ‘Nicer’ processes with a higher nice value will yield to more critical processes and will receive less CPU time. More information here.) Finally, if you’re using htop in a windowed environment with a mouse, you can use the mouse to select processes and options.

Another of the more interesting features of htop is the ability to tag multiple processes and then perform a single action on all tagged processes. So, for example, if you need to kill a group of several processes, you can tag all of the processes in the list with the space bar and then using the “F9 Kill” option. All of the tagged processes would then be killed. htop is a convenient, easy-to-use, and feature-rich replacement for top on your Linux system. It can be installed easily under Ubuntu with the following command:

sudo apt-get install htop

It can also be downloaded directly from the project website at http://htop.sourceforge.net.

If you have your own website, then getting your website to cooperate with search engines is critical to getting traffic to your site. There are many strategies for optimizing your website for Google, but optimizing Google for your website can also be beneficial. Google Webmaster Tools allows you to submit information to Google about your website and tweak how Google crawls and indexes your website. It also gives you feedback and statistics about how Google’s search bot sees your website. With Webmaster Tools you can be sure that Google is bringing the most amount of traffic to your website.

The first thing you can do with Webmaster Tools is to verify your website with Google. You can either insert a special meta tag within the HTML of your homepage which Google would then verify, or you can create a specific filename on your server. Once you’re verified, Google gives you access to the rest of their tools for your site.

In order to ensure that Google can get to every content page within your website, Google allows you to submit a sitemap of your website. A sitemap is an XML file which specifies page locations and update times for your website. Many server-side web applications like Wordpress can automatically generate a sitemap file for you. Once Google has the sitemap, it can more efficiently cover your entire website in their index.

If your website makes uses of the robots.txt file for restricting areas where you don’t want the bot to crawl, then Webmaster Tools has a feature for verifying that your robots.txt file is being properly interpreted by the Google crawl bot. Webmaster tools also gives you the ability to keep an eye on how high your Google PageRank is doing. It gives you an indicator that tells you what the PageRank distribution is across pages on your site.

Finally, Google Webmaster tools gives you an interesting look into exactly how your site is being indexed within Google. They give you charts that list all of the keywords that the crawler picked up when they crawled your site. They also give you tables indicating which search queries have had your content turn up in the search results, and how far down in the search results your site showed up. Also useful is the tool which shows you all of the other places on the Internet that Google found a link to your site. For each page within your site, Google Webmaster Tools lists every page on the Internet that links to that page. This is very helpful for discovering how much your site is interlinked from other sites and who is doing the linking. Also it allows you to see what kind of sites link to you and in what context they link to you. Finally, Google has a tool for summarizing what terms have been used in linking to your site. So, for example, if another site links to your site with a hyperlink that uses your full name then your full name will show up in the link text tool. These statistics are important for understanding what keywords are driving your page up in the search results.

There are several other features within Google Webmaster Tools which can help you better integrate your site with Google, but I’ve tried to cover just the main ones here. I’m sure that other search engines have tools similar to these, but since the majority of traffic to my website comes from Google, I’ve focused most of my energy on Google’s tools.

The following solution did not end up working for me. Azureus kept on crashing and erasing the log file didn’t really fix the problem. After more research, I discovered that the better solution is to download the original Jar file from here and uninstall the one from the Ubuntu repositories. I believe there is a bug filed for this under Ubuntu, so it should be fixed in future releases. The following is my original post in case you’re still interested.

I’ve been running the Azureus BitTorrent client on Ubuntu for a while. Recently it’s started having problems starting up. Usually it will be running and then it will crash spontaneously, and then when I restart it it loads until it gets to the main window and then immediately crashes. It will then crash every time I load it after that. Launching Azureus from the command line yielded a Java core dump error message. After searching around on the Internet and in forums, I found a tip that if you delete the files in ~/.azureus/logs, and then start it again it won’t crash. It works every time for me. The problem now is that this issue keeps popping up. A few days after fixing it, Azureus will crash again in the same fashion. So, here’s how you can modify Azureus to make sure that this issue gets taken care of automatically.

Before I continue, I should mention that I don’t have much experience with shell scripting. I just hacked around with this and it worked. If you need to keep your Azureus log files for some reason, then this hack probably isn’t for you. But I have never found these files to be useful and I assume most of the rest of you won’t be needing them either. The idea here is to modify the Azureus startup script so that it automatically cleans out your ~/.azureus/logs folder automatically every time it starts. Theoretically you should never have this crashing issue again. Or if it does crash, then it should automatically correct itself the next time you start Azureus. Here’s the drill:

From the command line, open the Azureus startup script:

gksudo gedit /usr/bin/azureus

Now find this line in the file (make sure you don’t delete this line):

exec java -Djava.library.path=/usr/lib/jni:/usr/lib \

And just above this line, add this line:

rm -r ~/.azureus/logs/*

This will tell the script to clean out the logs directory every time Azureus is started just before the actual application gets running. Now just save the file, and restart Azureus. Like I said before, shell scripting isn’t my forte, so if anybody else knows a better way to do this, feel free to comment.