Basic Linux Use
Using Linux via a command-line interface is not conceptually different from using an operating system that is mainly oriented toward a graphical user interface such as Microsoft Windows. The major difference is that on Windows you get the system and the software to do what you want by learning to juggle the activities of clicking left, right, or middle mouse buttons, scrolling the wheel, dragging or mouse around while holding one or more buttons on the keyboard or the mouse all while trying to precisely hit the many differently shaped and sized areas on the screen. On Linux, if you are using it as your desktop system, you can follow the same usage pattern you learned on Windows. However, when you log into the HPC system to get the job done via the command-line (CLI) interface you have to switch from juggling pictures to a more verbal mode and "talk to the system" using a language with a very small and well-defined vocabulary. The system will respond with either normal responses that show the results of your commands or with error responses that either mean that it didn't understand you or that the command you ran encountered a problem. All Linux commands have a help file that is called a "manual page" or manpage and can be called with the
man COMMAND command where COMMAND corresponds to the command you want to learn more about.
Think of a Linux system as a house with many rooms separated by the doors. And just as in a house getting where you need to be depends on where you start. As every house has an entrance every Linux system has its "root", which is marked by the topmost "door" -
"/". Every other "door" is also marked with
"/". The rooms with content are called files and storage rooms that hold other rooms are called directories and they all have names. So, if you know the name of the file or directory you need and the name of every directory between your target and the root of the filesystem you can always use an absolute path that starts from the root and ends up at the file or directory of your choosing. For instance, referencing a file
/project/bio/fasta/Ecoli_NC_008253.fa means you start from the root and go through the directories
project, bio, fasta to get to the room
Ecoli_NC_008253.fa that has what you need while opening the doors
/ between the rooms. This is called using an absolute path. If you think of all the paths on the system it quickly becomes apparent that together they create something similar to a tree as the rooms branch out and spread in many directions, but to go back up the tree there is only a single door to the parent directory. It's called a directory tree. If you are deep in the house i.e. the directory tree it can be cumbersome to travel from the entrance each time you need to pick up something in the same room where you already are or in an adjacent room. So, just as you would simply open the door to the parent room and then a door to the room you want you can do the same so called relative path traversal on Linux by using
".." as a symbol for the parental directory. So, if you are currently in the
"/project/bio/2bit" directory you can change to the
"/project/bio/fasta" directory or reference a file in that directory as
"../fasta/Ecoli_NC_008253.fa"</copy>. If for some reason you need to reference the directory you are in use
"." as a symbol for it as in
"touch ./my_file.txt". When you log into UFRC you are placed into your own room that's located at
"/home/USER" where USER is your username that you can see on the left side of the terminal prompt. By the way, we use
" quotes to make the commands stand out and it's not necessary to enter them at the command line.
- pwd - path to working directory - find out where you are right now.
- cd - change directory - go from one directory to another.
- ls - list - show the contents of the current working directory.
lsfor a short listing,
ls -ato show hidden files and directories (those with
.in front of their name as in .ssh,
ls -lto show a long listing akin to what you see in a detailed view in Windows Explorer or MacOS Finder. You can combine the switches as in
- touch - create an empty file or update a file or directory's access timestamp.
- mv [options] SOURCE DESTINATION - move or rename file(s) or directories.
mv test.txt test2.txt. The options are listed in the mv manpage (
- cp [options] SOURCE DESTINATION - copy file(s) or directories. The most useful option is -a, which copies the entire sub-tree. See the cp manpage for more details.
- rm [options] TARGET - remove - delete file(s) or directories. Useful options include -i for an interactive prompt that will give you a chance to reconsider your delete action that would be final otherwise (No going back!), -rf for a forceful removal of an entire directory tree, and '-v' to verbosely explain what's being done.
- file TARGET - learn some information about the target file.
Note: Linux is very liberal about you giving file and directory names, so by necessity it's very strict about referencing those names in commands. The test.txt and Test.txt are two different files and have to be referenced with the correct case. See the multitude of Linux tutorials and command cheat sheets for more details. For instance basic commands cheat sheet, moving around.
To be continued...