M01: Introduction to Operating Systems
 TU1: Installing, configuring and exploiting a computer system
 ASIX1
 Practical Exercise 9: Managing processes 2-2-12

Practical Exercise 9:  Managing processes

GENERAL CONDITIONS
1- Deadline: 16-02-2012.
2- Send your report as a PDF file attached to an e-mail with the following specifications:
     a) E-mail address: cf(at)collados.org or jordi.binefa(at)fje.edu depending who is your teacher
     b) File Name: asixdawA_surname_name_m01tu01pr9.pdf
     c) Subject: asixdawA_surname_name_m01tu01pr9.pdf
3- Make this report individually.
4- Left, right, top and bottom margins: 2cm.
5- Character format: a) Font:Times New Roman (or Liberation Serif), b) Size: 10, c) Questions typeface: Bold, d) Answers typeface: Regular
6- Page numbering on footer bar

1-  Basic Ideas
A program is only a set of code instructions and data stored in a file and it can be seen as a passive entity. A process is  a program  running,  and therefore, it can be seen as an entity in action, i.e, as a changing and dynamic entity. When the code and data of a program are sent to the RAM memory then the CPU can process the data using the instructions of the code. At that moment, a new process starts its execution.

During its lifetime, a process needs system resources such as the CPUs in the system to run its instructions or the system's physical memory or RAM to hold its data and code. Another kind of resources can be files in a storage device (hard disk, USB pendrive,....) or physical devices in the system such as a sound card. The allocation of proper resources to each process running in the system is a duty of Linux. Another duty of the operating system is to avoid that a single process monopolizes the most of the CPU time and RAM's space.

Linux is a multiprocessing operating system and therefore,  many processes are kept in memory at the same time. Usually there are more process running that CPUs in the system. Due to it, the processes  have to share  the CPUs time, being the operating system who decides what process is executed in each moment. The part of the operating system responsible for allocating CPU time to each proccess is called the scheduler or scheduling system. The scheduler is able to assign priorities to process. The higher  a process priority is, the higher will be its CPU time.

In Linux, four classes of processes can be distinguished:
a) Interactive processes: Interactive processes are started by someone connected to the system. These processes interact with their users, and therefore spend a lot of time waiting for keypresses and mouse operations. When input is received, the process must be woken up quickly, or the user will find the system to be unresponsive. Typical interactive programs are command shells, text editors, and graphical applications.
b) Daemons: These are server processes that run continuously. Most of the time, they are initialized at system startup and then wait in the background until their service is required. A typical example is the Apache web server. After the system is booted, the network daemon just sits and waits until a client program, such as Firefox, needs to connect.
c) Automatic (or Batch) procesess: These do not need user interaction, and  they often run in the background. Since such processes do not need to be very responsive, they are often penalized by the scheduler. Typical batch programs are programming language compilers, database search engines, and scientific computations.
d) Real-time processes: These have very strong scheduling requirements. Such processes should never be blocked by lower-priority processes, they should have a short response time and, most important, such response time should have a minimum variance. Typical real-time programs are video and sound applications, robot controllers, and programs that collect data from physical sensors.

In Linux, if you work with the shell, you can run a process in the:
a) Foreground:  The process occuppies the terminal that started the program, and you cannot start other applications as long as this process is running.
b) Background: The process is running but it does not occupy the terminal that started the program, so that terminal can accept new commands while the program is running. Running a process in the background is only useful for programs that don't need user input. Putting a job in the background is typically done when execution of a job is expected to take a long time.
The bash shell allows an easy handling of processes running in the foreground and the background thanks to a set of special commands. These commands switches processes between the foreground and the background.

In Linux, a process has a series of characteristics. The following are the most important:
a) The process ID or PID:  A unique identification number used to refer to the process. As long as the process exists, it keeps the same PID number. Different processes of the same program have different PIDs. When a process is ended, its PID is freed and eventually the operating system can assign this number to another process.
b) The parent process ID or PPID: the number of the process (PID) that started this process.
c) Nice number: It is the degree of friendliness or niceness of this process toward other processes. This number is closely related to process priority. The higher the nice number is, the lower the priority of the process is.
d) Real User  ID or RUID: The owner of the process, i.e., the user  issuing the command.
e) Effective User  ID or EUID: The one determining access and permissions to system resources. RUID and EUID are usually the same user.


Linux processes have the following states:
a) Running: The process is either running (it is the current process in the system) or it is ready to run (it is waiting to be assigned CPU time).
b) Waiting: The process is waiting for an event or for a resource. Linux differentiates between two types of waiting process; interruptible and uninterruptible. Interruptible waiting processes can be interrupted by signals whereas uninterruptible waiting processes are waiting directly on hardware conditions and cannot be interrupted under any circumstances.
c) Stopped: The process has been stopped. At the moment, no CPU time will be assigned to the process.
d) Zombie: An ended process, which has not freed its PID for some unknown reason. In fact, the process no longer exists, and there are no system resources allocated for this process, but its PID can not be freed. Commands such as ps, which shows a list of processes, will display the process even if it no longer exists. The most of the times, a zombie process is due to a programming error.

In the present practical exercise we are going to work with a set of commands and GUI programs which allow us to display and manage the process running in our system. We will work with the following commands: ps, pstree, top, kill, killall, nice, renice, jobs, fg and bg. Additionally, we will work with the GNOME System Monitor, a useful GUI program.

 2-Displaying running processes

2.1-  ps command
Display information about the active processes.  It is a complex command with a wide range of options and parameters. Particularly interesting are the following two modes of executing this command: Inf
a) ps  aux  --> This command displaying basic information (euid, pid, %memory, %cpu, stat, command, ....) about every process running in the system.
b) ps  -eo  pid,ppid,euid,euser,ruid,ruser,ni,%cpu,%mem,stat,start,time,args  -->  This command displaying  information about the process ID, parent process ID, effective and real user (number and name), % of CPU time and % of RAM allocated to the process, nice number, state of the process, what time the command was started, cumulative CPU time and command name (with arguments). This command displays information about every process running in the system.

The next example shows what happens when the  ps  aux  command is executed:



The most important columns are:

USER: Effective user's name
PID: Process ID.
%CPU: % of CPU time
%MEM: % RAM memory
STAT: State (R = Running S=Sleeping T=Stopped, Z=Zombie)
COMMAND: Command (with arguments) executed.

2.2- pstree command
Display a tree of processes. Using this command, the parent-child relationships of processes is shown on the screen. The next example shows what happens when the  pstree  command is executed:



The command pstree -up displays the tree of processes and information about the PID and UID of each process.
2.3- top command
The program provides a dynamic list in real-time of all running processes. Information about  the process ID, user ID, % of CPU time and % of RAM allocated to the process, nice number, state of the process, what time the command was started and command name (with arguments) will be show on the screen. The next example shows what happens when the  top  command is executed:



Press the p key if you want to end the top command and come back to the terminal.
 
2.4- Working with the GNOME System Monitor
The tab "Processes" of GNOME System Monitor (System --> Administration --> System Monitor) displays information about the active processes in a friendly GUI enviroment:



IMPORTANT NOTES:
a) Select  the All Processes option of the View menu.
b) Select the following options in Information Fields of Edit --> Preferences --> Processes: Process name, User, Status, %CPU, Started, Nice, ID, Command Line and Memory.


3- Ending processes
3.1- kill command and signals
Processes end because they receive a signal. There are multiple signals that you can send to a process. If you want to send a signal to a process, you will have to use the kill command. The command kill -l shows a list of signals. Some of the most important signals, are the following:
Signal nameSignal numberMeaning
SIGTERM15Terminate the process in an orderly way (By default, if no signal is written, the kill command send the SIGTERM signal to a process).
SIGINT2Interrupt the process. A process can ignore this signal.
SIGKILL9Interrupt the process. A process can not ignore this signal.
SIGHUP1For daemons: reread the configuration file.

To kill a process you need to know its PID. The ps command can provide you this information.

The  synopsis  of  the kill command is :  kill    <-SIGNAL>  PID. 

The next example shows how we can run  the  kill command to end a process using its PID:



Now, if we run the ps command, the the next information will be displayed on the screen:



Some additional examples:
a) kill  -SIGTERM  4388 (equivalent to kill  -15  4388 or kill  4388)
b) kill  -SIGKILL    4388 (equivalent to kill  -9  4388)

As SIGKILL is a method of last resort and inherently dangerous (data used by the process can be definitely lost), you should use this signal cautiously.  You should send the termination signal SIGTERM when you need to end a process. Only if this signal does not work should you attempt to use the -SIGKILL signal.

3.2- killall
It is an alternative to command kill. The killall command kills all processes associated with programs whose names are provided to it as arguments. It is an alternative to command kill. The killall command kills all processes associated with programs whose names are provided to it as arguments. It is an alternative to command kill.  A signal may be provided as an argument as well. If no signal is specified, SIGTERM is sent.

The  basic synopsis  of  the kill command is :  killall    <-SIGNAL>  program_name(s).  Signals can be specified either by name (e.g.  -SIGKILL ) or by number (e.g. -9).  A killall process never kills itself (but may kill other killall processes).

The next example shows how we can run  the  killall command to end all instances of the geany text editor:


The killall command can end two or more process associated with program if a list of program names is provided to killall as arguments. For instance:
killall geany firefox-bin nano

3.3- Killing processes with the GNOME System Monitor
If you want to end a process using the GNOME System Monitor, you have to select the process (firefox-bin in the next example)  and then press the virtual key "End Process" at the right bottom corner. Example:



4- Process priority: nice and renice commands
Linux can run multiple processes. It does this by sharing the CPU and other resources among the processes. When you only have one or a limited number of CPUs, you need to decide how to share those limited CPU resources among several competing processes. This is generally done by selecting one process for execution and letting it run for a short period, or until it needs to wait for some event, such as IO to complete. To ensure that important processes get enough CPU time, a selection must be done based on a scale of "priorities". By using this scale we can allocate our CPU resources more appropriately.  High priority programs like daemons, services or I/O processes can be set to receive more of the CPU’s focus, while lower priority programs, which are not so important, can be set to receive a lesser CPU’s focus.

In Linux, processes have a priority number called "nice value" , which is a number between -20 and 19. The value of -20 is the highest, and 19 is the lowest priority with a default of zero. Process priority can be set with the nice command and changed using the renice command. Larger nice values correspond to a lower priority (you are being nice to the other processes on the system). Processes with a lower nice value (higher priority) run before processes with a higher nice value (lower priority). The nice value also determines the CPU time assigned to each process. 

4.1- Displaying the nice value
If you wnat to display the priority of processes running in you system, you can run one of the following commnads:
a) ps  -eo  pid,user,ni,%cpu,%mem,stat,start,time,args
b) ps -alx

4.2- Starting a process with a different priority (not its default priority). Using the nice command.
The nice  command (with -n option) is used to start a process with a different priority. Keep in mind the following rules:
a) -n option with a positive value ==>  The nice values is increased ==> The priority of your process is descreased.
b) -n option with a negative value ==>  The nice values is decreased ==> The priority of your process is increased.
c) You usually have to be root to specify negative value.
Examples:
a) Starting, as a regular user,  the geany program with a lower priority (or higher niceness):
nice  -n  5  geany &
Geany has been started with a decrease in its default prioriy equal to 5.

b) Starting, as root user,  the geany program with a higher priority (or lower niceness):
nice  -n  -4  geany &
Geany has been started with an increase in its default prioriy equal to 4.

4.3- Changing priority of a running process. Using the renice command.
The renice  is used to change the priority of a running command. Keep in mind the following rules:
a) Opposite to nice, renice works with absolute priority positive value. The written value written will be the value assigned directly to the process' nice value.
b) You have to be root to run the renice command and decrease a nice value below its default value.
c) You will need the PID.
 Examples:
a) Changing the priority of process 6876 :
renice  +10  6876
The nice value of process 6876 has been changed to 10.

a) Changing the priority of process 9854 :
renice  -15  9854
The nice value of process 9854 has been changed to -15.

4.5- Using the GNOME System Monitor

You can use GNOME System monitor to change the priority of the process. Launch System Monitor from System > Administration > System Monitor and go to the Processes tab as shown in the following screenshot:

Now, you will be able to see the Nice column in front of every process. It basically specifies the priority value of the process. You can change priority  of any process by right clicking the Nice column value and choosing Change Priority as shown in the following screenshot.

Thats all, adjust the new priority value and click Change Priority.


PRACTICAL EXERCISE

1- Open the terminal. Find out its PID and name using the ps command and the  GNOME System Monitor. Is there any difference between PID given by the command and the GUI program?. Why?
2- Run a command to show only the relevant information about the terminal process. What is the problem with this command?. How would you avoid that problem?
3- End the terminal process using the GNOME System Monitor.
4- Open the terminal. Find out its PID. Compare its current PID and the PID found in the 1st question. Is there any difference between them? Why?.
5- End the terminal process using the kill command.
6- Start the OpenOffice Word  Processor.  Using the ps command, show information about  the process ID, effective user (number and name), % of CPU time and % of RAM allocated to the process and command name (with arguments). Write some text in the new document. Do not save the text. End the process using the kill command. Start the OpenOffice Word process and the System recovery process. Have you recovered the text written? Why?.
7- Open the OpenOffice Word Processor. Open a termional and change to user fje. As fje user, find out the PID and user ID of the OpenOffice Word Processor.  Try to kill the process. Are you able to kill it? Why?.
8- Become another time your "by default" user. Change to root. As root user, find out the PID and user ID of the OpenOffice Word Processor.  Try to kill the process. Are you able to kill it? Why?.
9- As root user, find out the PID and user ID of the cups server.  End the cups server using the command kill and the signal SIGHUP.  Is there any change in the PID process?. Why?.
10- End the cups server using the command kill and the signak SIGTERM. Check if the process has been ended.
11- Start three instances of geany text editor and two new instances of GNOME system editor. End these 5 processes using their names and one only command. Check the result of your command.
12- Start geany. Working with the terminal and as a regular user (your  user by default) check its nice value by default. Check the nice value of geany using the GNOME System Monitor. Is there any difference between the nice value given by the command and the GUI program?. Why?.
13- As a regular user, start a new instance of geany with an increase in its default nice value equal to 7. Check if geany is running with the new nice value. Have you had any problem? Why?.
14- As a regular user, start a new instance of geany with a decrease in its default nice value equal to 6. Check if geany is running with the new nice value. Have you had any problem? Why?.
15- As root user , start a new instance of geany with a decrease in its default nice value equal to 6. Check if geany is running with the new nice value. Have you had any problem? Why?.
16- As root user, close all instances of the geany program.
17- As a regular user, start a new instance of geany with an increase in its default value equal to 3. Now, try to change its nice value to -10. Are you able to change the nice value of geany? Why?.
18- As a regular user, start a new instance of geany with an increase in its default value equal to 7. Now, try to change its nice value to 2. Are you able to change the nice value of geany? Why?.
19- As a root user, change the nice value of the process geany  started in the previous question. Try to change its nice value to -13. Are you able to change the nice value of geany? Why?.
20- As a regular user, start a new instance of geany with an increase in its default value equal to 7. Now, try to change its nice value to 2 using the GNOME Monitor System. Are you able to change the nice value of geany? Why?.
21- Now, try to change its nice value to -4 using the GNOME Monitor System. Are you able to change the nice value of geany? Why?.
22- Find out two different ways of changing the nice value using the GNOME Monitor System.