[ previous ] [ Contents ] [ 1 ] [ 2 ] [ 3 ] [ 4 ] [ 5 ] [ 6 ] [ 7 ] [ 8 ] [ 9 ] [ 10 ] [ 11 ] [ 12 ] [ 13 ] [ A ] [ B ] [ C ] [ D ] [ next ]
To keep your Debian system running smoothly, you need to do some routine tasks. These tasks usually need to be done more frequently on a network, but they are also important on standalone desktop machines.
The basic administration tasks are:
Managing Users.
Managing Groups.
Managing Printing and Print Jobs.
Managing Processes.
Backing up and Compressing Files.
Scheduling Tasks.
Checking Filesystems.
Only the root user can do all these tasks. Regular users can only do some of these tasks for their own accounts.
These tasks are generally done from the command line. Although desktop tools exist for some of them, few are as useful or as versatile as the commands that they are based on.
Linux is designed for multiple users. To use the system, you must log in with a user account. A user account consists of a user name, a password, and a number of saved settings.
User accounts are organized into groups of accounts with similar access to the system. The software and the hardware you can use depends on your user accounts and the groups that it is in. See Changing Individual Passwords, Section 12.1.4.
New users can be added by the root user. Setting up a new user means:
Adding the user to the list in /etc/passwd.
Assigning an initial password. The user may change this password later.
Creating a home directory, a mail file, and startup files.
While you can configure files in a text editor, the easiest way to add a user is to use the useradd and passwd tools.
Log in as the root user. Then:
1 Create the new user. Type: useradd options user . The new user is added to the end of the /etc/passwd file.
[Note]
To fully set up a typical new account, type useradd -mk user .
2 Create an initial password. Type: passwd user . The password is now added to the /etc/shadow file.
The useradd Command (Sets Up New User)
+-------------------------------------------------------------------+
| Option | Action | Comments |
|---------+-----------------------+---------------------------------|
| |Adds a home directory | |
| |that: | |
| | | |
| | * Does not the same |If you do not want a home |
| | name as the user |directory with these features, |
|-d | name. |use the -m option . |
|directory| | |
| | and/or |If -d or -m is not used, then no |
| | |home directory is created. |
| | * Is not a | |
| | sub-directory of / | |
| | home. | |
|---------+-----------------------+---------------------------------|
|-e date |Adds an expiry date for| |
| |the user account | |
|---------+-----------------------+---------------------------------|
| | |Enter multiple groups by name or |
| |Adds groups in addition|group id ( gid) . |
|-G groups|to the default one. | |
| | |Separate the names with a comma |
| | |and leave no blank spaces. |
|---------+-----------------------+---------------------------------|
| |Copies default files to|Use with the -m or -d option. |
|-k |the user's home | |
|directory|directory. |By default, files are copied from|
| | |/etc/skel. |
|---------+-----------------------+---------------------------------|
| | |The default features are: |
| | | |
| | | * That has the same name as the|
| | | user name. |
| |Creates a home | |
|-m |directory with the | and |
| |default features. | |
| | | * Is a sub-directory of /home. |
| | | |
| | |If you do not want both these |
| | |features, then use the -d option.|
|---------+-----------------------+---------------------------------|
| |Disables the account |Use with the -e option, which |
|-f number|after the number of |sets an expiry date. |
| |days has passed. | |
|---------+-----------------------+---------------------------------|
| |Changes the default |If no group is specified, then a |
|-g group |group. |group with the same name as the |
| | |user name is created. |
|---------+-----------------------+---------------------------------|
|-s shell |Sets the users' shell. |If no shell is specified, then |
| | |the default bash is used. |
+-------------------------------------------------------------------+
1 Log in as the root user.
2 Type: userdel -r user .
[Warning]
The -r option removes the user's home directory and the files in it. If you want to keep the home directory, omit the -r option.
Log in as the root user. Then type: usermod options user .
The options are the features to edit. The options are similar to those of the useradd command.
The usermod Command (Edits User Accounts)
+-------------------------------------------------------------------+
| Option | Action | Comment |
|---------+-------------------------+-------------------------------|
| | |Enter multiple groups by name |
|-G groups|Adds groups in addition |or gid . Separate the names |
| |to the initial one. |with a comma and leave no blank|
| | |spaces. |
|---------+-------------------------+-------------------------------|
| | |Enter multiple groups by name |
|-g group |Changes the default |or gid . Separate the names |
| |group. |with a comma and leave no blank|
| | |spaces. |
|---------+-------------------------+-------------------------------|
|-d |Changes the home | |
|directory|directory. | |
|---------+-------------------------+-------------------------------|
|-l new_ |Changes the user name. |The user must be logged out |
|name | |when this option is used. |
|---------+-------------------------+-------------------------------|
|-e date |Sets an expiry date for | |
| |the user account. | |
|---------+-------------------------+-------------------------------|
| |Sets the number of days | |
|-f number|after the expiry date |Use with the -e option, which |
| |that the account is |sets an expiry date. |
| |disabled. | |
|---------+-------------------------+-------------------------------|
| | |The user should not be |
| | |executing any command when this|
| | |option is used. |
| | | |
|-u uid |Sets the user id . |Must be a unique id . |
| | | |
| | |Files in the user's home |
| | |directory are changed, but not |
| | |files elsewhere. |
|---------+-------------------------+-------------------------------|
|-s shell |Sets the users' shell. |If no shell is specified, then |
| | |the default bash is used. |
+-------------------------------------------------------------------+
1 Type passwd user . You are prompted for (current) UNIX password
You do not need to enter the user for the current account. For information about selecting a password,
2 Enter your current password. You are prompted for your New UNIX password .
3 Enter your new password. You are prompted to retype new UNIX password .
4 Enter your new password again. A message displays: All authentication tokens updated successfully.
The next time you log in to the account whose password you have changed, you must use the new password.
A good password is one that is hard to crack. That is, it is:
A mixture of letters, numbers, symbols, and of upper and lower case.
At least six characters long.
Not based on anything associated with you, such as a birthday, a spouse's name, a pet, or a favorite sports team or movie.
Not a common word found in a dictionary.
Different from your passwords on other machines.
Different to your user account.
Not written down anywhere.
User groups can define:
How users can access the system.
What files, directories, and services uses can work with.
What hardware users can work with.
Some of these features are defined when the group is created. Others are created by using permissions or configuration settings.
Groups are listed in the /etc/group file. Each group has a unique id ( gid ).
The /etc/group file lists all existing groups. By default, any user can view this file, but you must be logged in as the root user to edit it.
Each group is defined in a line in /etc/group. Columns are separated by a colon. Information is given in the following order:
group: encrypted password: group_id: users_in_group
Note:
The group name generally reflects what access the group has to the system. For example, the lp group defines who can use the printer.
The encrypted password is rarely used any more. Usually, an x is added as a placeholder.
The group id is usually a number from 1 to 50.
The list of users in the group only needs to contain those who are not in the group by default.
You can edit /etc/group directly in a text editor. However, often, the command line tools are easier to use.
1 Log in as root user.
2 Type: groupadd -g gid group .
You can create a group without a group id by leaving out the option.
1 Delete the group. Logged in as root user, type: groupdel group
2 Find files that belong to the group. Type: find -l -group group Searching for Files and Text, Chapter 8
3 Change the group ownership on the files. Managing Groups, Section 12.2
To save effort, select multiple files.
1 Select the group. Logged in as root user, type: groupmod -g gid
2 Find files that belong to the group. Type: find -l -group group Searching for Files and Text, Chapter 8
3 Change the group ownership on the files. Managing Groups, Section 12.2
To save effort, select multiple files.
Linux includes a number of command line tools for administering printing. These commands are useful on networks, or for troubleshooting and managing printing on any system.
Each printer on a Linux system has its own spool or queue in the /var /spool/lpd directory. This spool uses the short name of the printer, which is created when the printer is first configured. When a file is printed, it becomes a job in the spool. Each job is stored in a spool until it is printed.
Printing tools generally start with lp (short for "line printer"). If you are familiar with other Linux commands, you can usually make a good guess at what the rest of the command should be. For example, lprm removes a print job from the spool
You can use the following tools for managing printing:
lpq : Checks the status of print jobs.
lprm : Removes a print job from a spool.
lpstat : Reads the status of a spool.
Type: lpq options user .
The command's name comes from the fact that a spool is sometimes called a "queue."
The lpq Command (Print Jobs)
+-------------------------------------------------------------------+
|Option | Action | Comments |
|-------+----------------------------------+------------------------|
|-l |Prints information for each file | |
| |in the printer queue. | |
|-------+----------------------------------+------------------------|
|number |Checks the status of a job number.| |
|-------+----------------------------------+------------------------|
|-P |Sets the printer that a job should|Not needed to use the |
|printer|use. |default printer. |
+-------------------------------------------------------------------+
Type: lprm options job user .
The command's name comes from the fact that it removes (rm) a job from a line printer (lp).
Only the root user can remove a job sent by another user.
Regular users can remove their own print jobs.
Use lpstat to find job numbers. See Reading the Status of a Printer Spool, Section 12.3.3.
The lprm Command (Remove Print Jobs)
+-------------------------------------------------------------------+
|Option | Action | Comments |
|-------+---------------------------------+-------------------------|
|-P |Sets the printer to use. |Not needed if the default|
|printer| |is used. |
|-------+---------------------------------+-------------------------|
|- user |Removes all jobs sent by the user| |
| |specified. | |
+-------------------------------------------------------------------+
Type: lpstat -t options .
Type: lpr options files .
The lpr Command (Print)
+-------------------------------------------------------------------+
|Option | Action | Comments |
|-------+-----------------------+-----------------------------------|
|-P |Selects the printer to |Otherwise, the system default is |
|printer|use. |used. |
|-------+-----------------------+-----------------------------------|
| |Sends a symbolic link |This option saves time and disk |
|-s |to the printer instead |space for large files, but may |
| |of the file. |cause printing to be slower. |
+-------------------------------------------------------------------+
A process is a running program to which the Linux kernel has assigned a share of your system's resources or computing power. A process that starts when the system starts and runs in the backgrounduntil needed is a daemon. Because Linux is a multi-tasking operating system, dozens of processes of all sorts can be running at any one time.
Some of the reasons that you may want to check the processes that are running include:
To Troubleshoot: Your system is not responding properly. The processes, the combination of processes, and the resources used by processes should all be examined as a routine part of troubleshooting.
To check system security: If you suspect an intruder on the system, the processes used may provide a clue. You should also check what user or group id a process is running under. In particular, check programs running by the root user and see if anything is out of the ordinary.
To close a frozen or strangely behaving program: While Linux itself is stable, the desktop has developed rapidly, and some programs that run under the X Window System behave erratically. You may also need to close a program while testing it.
Linux uses three basic types of processes:
Interactive: An ordinary program started by a user. Many commands entered from the command line are interactive processes. So is each new window opened in the X Window system.
Batch: Any process run without a command line terminal.
Daemon: A process that runs continuously. Usually, they are started when the system starts, but they can also be started by the root user. Daemons generally run in the background.
You should also know:
A major characteristic of each process is its priority. If system resources are short, the priority determines what share of the resources the process uses. In extreme cases of resource shortages, the priority determines the order in which processes can use system resources.
Processes can be hierarchal. For example, the printing daemon is always running, but a new copy of it is started for an individual print job, then closed after the job is printed. The same thing happens when you open a command line, then enter a command. In both cases, the original process is called the parent, and the secondary process is called the child. Because of this type of relation, you may sometimes hear about process trees.
The opening of a secondary process is also called forking.
Linux includes a number of tools for working with processes. See Viewing Intensive Processes, Section 12.4.4.
[Warning]
Be careful about killing a process or changing its priority. Unless you know what you are doing, you can easily crash your system. Be especially careful of daemons, which often manage important system services and resources.
Understanding Process Statistics
+-------------------------------------------------------------------+
|Statistics| What It Means |
|----------+--------------------------------------------------------|
|command |The command used to start the process |
|----------+--------------------------------------------------------|
|%cpu |The percentage of processing power used by the process. |
|----------+--------------------------------------------------------|
|% mem |The percentage of system memory used by the process. |
|----------+--------------------------------------------------------|
| |The process's nice scheduling priority. The higher the |
|ni |number, the lower the priority. |
| | |
| |Negative numbers are highest of all. |
|----------+--------------------------------------------------------|
|pid |Process id . Useful for killing a process. |
|----------+--------------------------------------------------------|
| |The process's scheduling priority. The higher the |
|pri |number, the lower the priority. |
| | |
| |Negative numbers are highest of all. |
|----------+--------------------------------------------------------|
|rss |The amount of ram memory in kilobytes. |
|----------+--------------------------------------------------------|
|shared |The amount of shared memory used. |
|----------+--------------------------------------------------------|
|size |The size of the virtual image. |
|----------+--------------------------------------------------------|
|start |The date that the process was started. |
|----------+--------------------------------------------------------|
| |The state that the process is in. Common states are: |
| | |
| | * S: Sleep. |
|stat | * T: Suspended, but running. |
| | * R: Running |
| | * D: Asleep and not interruptible. |
| | * Z: Zombie |
|----------+--------------------------------------------------------|
|swap |The amount of swap memory used, in kilobytes. |
|----------+--------------------------------------------------------|
|time |How much cpu time the process is using. |
|----------+--------------------------------------------------------|
|tty |Controlling Terminal. Daemons do not have a terminal. |
|----------+--------------------------------------------------------|
|user |The user running the process |
|----------+--------------------------------------------------------|
|wchan |The kernel function which controls the process. |
+-------------------------------------------------------------------+
1 Type: gtop . The gnome System Monitor opens.
2 To view the details of a process in a separate window, right-click on it and select Details from the pop-up menu.
3 To select the characteristics to display, select from the menu Settings > Preferences > Process Fields.
Type: ps options .
[Note]
Unlike most commands, ps does not require one or two hyphens before the options.
For a complete view, you must be root user.
Once the display is running, you can use the O+ or O- options to add or subtract statistics from the display. The display you set by the keyboard is used in the terminal in which you are running the command as long as the terminal is open, or until you change it.
The ps Command (Process Screen)
+-------------------------------------------------------------------+
|Option | Action | Comment |
|-------+--------------------+--------------------------------------|
|l |Displays in long | |
| |format | |
|-------+--------------------+--------------------------------------|
|u |Displays user name | |
| |and start time | |
|-------+--------------------+--------------------------------------|
| |Does not display of | |
|r |processes that are | |
| |not running | |
|-------+--------------------+--------------------------------------|
| | |A plus sign adds a column the next |
| * O +| |time you use the command in the same |
| key| |terminal. |
| | | |
| or |Adjusts displays |A minus sign makes the column not |
| | |display the next time that the command|
| * O -| |is used in the same terminal. |
| key| | |
| | |See Display Options for the ps |
| | |Command below. |
+-------------------------------------------------------------------+
Display Options for the ps Command
+-------------------------------------------------------------------+
| Option | To display | Comment |
|---------+-------------------------------+-------------------------|
| * k |The cpu time taken by the | |
| * stime|process | |
|---------+-------------------------------+-------------------------|
| * U |The id of the process's user | |
| * uid | | |
|---------+-------------------------------+-------------------------|
| * c |The name of the processes' | |
| * cmd |binary file | |
|---------+-------------------------------+-------------------------|
| * g |The process's group | |
| * pgrp | | |
|---------+-------------------------------+-------------------------|
| * p |The process's id | |
| * if | | |
|---------+-------------------------------+-------------------------|
| * p |The process's parent's id | |
| * ppid | | |
|---------+-------------------------------+-------------------------|
| * s |The process's size |The size is given in |
| * size | |kilobytes |
+-------------------------------------------------------------------+
The top command shows a constantly updated summary of the processes that are using the most system resources. The figures given are relative to each other, rather than representing the actual amount of ram used.
Type: top options .
Once the display is running, you can use keyboard commands to:
Change the details of the display, including the time between updates and the statistics shown.
Change the process or number of processes shown.
Sort the display according to different criteria.
Set a priority, as with the nice command. See Setting the Priority for a Process From the Command Line, Section 12.4.8.
Stop a process as with the kill command. See Stopping a Process From the Desktop, Section 12.4.5
The top Command (Views Intensive Processes)
+---------------------------------------------------------------+
| Option | Action |
|------------+--------------------------------------------------|
|-d seconds |To set the interval between updatings of the |
| |display. |
|------------+--------------------------------------------------|
|-p |To display only processes with a particular id . |
|process_id | |
+---------------------------------------------------------------+
Keyboard Shortcuts for the top Command
+---------------------------------------------------------------+
|Press.| To. . . |
| . . | |
|------+--------------------------------------------------------|
|c |Toggle the display of the command name or full command |
| |line. |
|------+--------------------------------------------------------|
|f |Select fields to display. |
|------+--------------------------------------------------------|
| |Open a prompt to enter a process id to kill and send a |
|k |signal to kill it. Unless you are the root user, you can|
| |only kill processes you started. |
|------+--------------------------------------------------------|
|l |Toggle the display of load average and uptime |
| |information. |
|------+--------------------------------------------------------|
|m |Toggle the display of memory information. |
|------+--------------------------------------------------------|
|M |Sort by the use of resident memory. |
|------+--------------------------------------------------------|
|N |Sort by process id . |
|------+--------------------------------------------------------|
|n |Set the number of processes to show. Selecting 0 |
|number|displays a screenful. |
|------+--------------------------------------------------------|
|q |Exit the display. |
|------+--------------------------------------------------------|
| |Set priority, as with the nice command. |
|r | |
| |See Setting the Priority for a Process From the Command |
| |Line. (***) |
|------+--------------------------------------------------------|
|space |Update the display. |
|bar | |
|------+--------------------------------------------------------|
|t |Toggle display of processes and cpu states. |
+---------------------------------------------------------------+
***) See Setting the Priority for a Process From the Command Line, Section 12.4.8.
1 Type: gtop. The gnome System Monitor opens.
2 Select a process. A process is highlighted when selected.
3 Right-click on the process and select Kill Nicely ( Sigterm ) from the pop-up menu. The process is killed without any confirmation dialog.
[Warning]
Closing some proceses can crash your system or cause other problems.
You can stop a process from the command line in three ways:
Identifying it by id .
Identifying it by the command that starts it.
Identifying it by a file that uses it.
Type: kill process_ids .
Unless you are root user, you can only kill processes that you started.
[Warning]
Closing a process can crash your system or cause other problems.
Type: killall options command .
Unless you are root user, you can only kill processes that you started.
[Warning]
Closing some processes can crash your system or cause other problems.
Type: killall options file .
A name with a forward slash in it is read as a file name. All processes that use that file are closed.
Unless you are root user, you can only kill processes that you started.
[Warning]
Closing some processes can crash your system or cause other problems.
The killall Command (Stops Processes)
+-------------------------------------------------------------------+
|Option| Action | Comment |
|------+-------------------------+----------------------------------|
| |Stops only processes that|If this option is not used, all |
|-e |are an exact match of |processes are killed that match |
| |what you enter. |the first 15 characters. |
|------+-------------------------+----------------------------------|
|-g |Stops all parent | |
| |processes as well. | |
|------+-------------------------+----------------------------------|
| |Enters a confirmation | |
|-i |before each process is | |
| |stopped. | |
|------+-------------------------+----------------------------------|
|-v |Displays progress | |
| |on-screen. | |
+-------------------------------------------------------------------+
Linux assigns priorities for processes. However, you can set your own priorities:
1 Type: gtop . The gnome System Monitor opens.
2 Select a process. The process is highlighted when selected.
3 Right-click on the process and select Renice from the pop-up menu. The Renice window opens.
4 Drag the slider bar with the mouse cursor to set the priority. The priority is a number with a range from -20 (highest priority) to 19 (lowest priority). Only the root user can use a negative priority (-1 to -20).
5 Select the Apply button to change the priority, then the ok button to close the window.
Linux assigns priorities for processes. However, you can set your own priorities.
Type: nice -n adjustment command .
In this structure:
The adjustment is a number with a range from -20 (highest priority) to 19 (lowest priority). Only the root user can use a negative priority (-1 to -20).
The command includes any options or arguments.
[Warning]
Changing a process's priority may cause problems on your system, especially on systems with limited hardware resources or a large number of users.
Regular back ups are a basic precaution against lost of date due to security leaks, power loss, or user carelessness.
[Note]
Use back up commands with crontab to have automatic, regularly scheduled backups. See Backing Up and Compressing Files, Section 12.5.
Debian supports most backup devices that use ide , scsi or usb standards, or use a parallel port. These devices include tape, zip and iomega drives, as well as removable hard drives. However, for some of these devices, you may need to install a package that is not installed by default. Be prepared to do some research before backing up to such devices.
Before doing a backup, plan it carefully. Consider:
Which files are irreplaceable without a back up. Irreplaceable files probably include those in users' home directories (including /root), and configuration files, such as those in the /etc/ directory.
Which files are on removable drives, such as cd s or floppies. Since you probably do not need to back up removable drives, you might unmount them before doing a complete system backup.
Which files can be easily replaced by installing a package or doing a selective install or upgrade of the operating system. You can save time and storage space by not including these files in a backup.
Which files are unnecessary or dangerous to backup. For example, files in /tmp are probably unnecessary, while restoring some files that are in the /proc directory could crash the system.
Whether to compress files using gzip or bzip2 . Compressing saves space, but adds another step to the backup. Also, while compression is generally reliable, it creates another stage at which the process can fail.
Whether users are responsible for backing up their own files. Since only the root user has full permissions for all files on the system, usually backups are best done by the root user. However, if users back up their own files, you might omit backing up the home directory, or at least not back it up regularly.
Linux has several tools for backing up and restoring files:
dump / restore : Old tools that work with filesystems, rather than files, and can back up unmounted devices. Although you can easyly control what is backed up with dump by editing a single column in the /etc/fstab file, for some reason these utilities have fallen into disuse. Today, many distributions of Linux, including Debian, do not even include them by default. If you want to use dump and restore , you must install them yourself.
tar : A standard backup tool, and by far the easiest to use. It is especially useful for backing up over multiple removable devices using the -M option. See Archiving and Compressing From the Desktop, Section 12.7.
cpio : A very flexible command, but one that is hard to use because of the unusual way in which the command must be entered.
Many commercial or free software back up tools are also available.
The cpio command is one of the most commonly used Linux back up tools.
The cpio command has two unusual features:
Unlike tar , in which the files to back up are typed in as part of the command, cpio reads the files to work with from the standard input (in other words, the screen).
This feature means that cpio must be used as part of a multiple command or with a redirection pipe. Examples of this usage are shown in the tables below.
cpio must always be used with one of three flags. Flags are options that set the mode in which the command runs. Only one flag can be used at a time, and it must come before any other options. In addition, the choice of flags limits the options that can be used. Each flag also has a gnu option that can used in its place. The gnu option gives a convenient name for each flag: extract, create, and pass- through.
The cpio Command (Backup and Restore)
+---------------------------------------------------------------+
| Flag | GNU Option | What it does |
|---------+--------------+--------------------------------------|
| | |Restores files from an archive that |
| | |match the string. |
| | | |
|-i | |Metacharacters can be used in the |
|options |--extract |string. See Using Metacharacters or |
|string | |Wild Cards. (***) |
| | | |
| | |Strings should be surrounded by double|
| | |quotation marks. |
|---------+--------------+--------------------------------------|
|-o |--create |Backs up files to an archive file. |
|options | | |
|---------+--------------+--------------------------------------|
| | |Copies files to another directory on |
|-p | |the same system. |
|options |--pass-through| |
|directory| |This option is useful for repeating a |
| | |backup that is later burned to a cd . |
+---------------------------------------------------------------+
***) See Using Metacharacters or Wild Cards, Section 5.3.6.
To do a backup, use cpio with a search command, such as find . See Searching for Files: Using the find Command, Section 8.4.1.
The basic structure is: find -name string -print | cpio -o options > directory .
In this example:
The -name option for find lets you search for a string enclosed in double quotation marks. Metacharacters can be used. See Using Metacharacters or Wild Cards, Section 5.3.6. The bar character ( | ) redirects the output of find to cpio .
The -o flag sets cpio to create an archive file for backing up.
The target is a directory.
The > redirection operator redirects files to the location for the back up. Typically, this location is on a removable device.
The -o Flag in the cpio Command (Backup)
+---------------------------------------------------------+
| Option | Action |Comments|
|-------------+----------------------------------+--------|
| * -A |Adds files to an existing archive.| |
| * --append | | |
|-------------+----------------------------------+--------|
| * -v |Displays progress on-screen. | |
| * --verbose| | |
+---------------------------------------------------------+
To restore files from an archive file, use cpio with a redirection pipe. Type: cpio -i options string < target .
In this example:
The -i flag sets cpio to restore files.
The string describes the files to restore. Metacharacters can be used. See Using Metacharacters or Wild Cards, Section 5.3.6.
The < pipe redirects the files described by the string to the target.
The target is the directory to which to restore files.
The -i Flag for the cpio Command (Restore)
+---------------------------------------------------------------+
| Option | Action | Comments |
|----------------------+-------------------+--------------------|
| * -f |Copies all files | |
| * --non-matching |except those that | |
| |match the string. | |
|----------------------+-------------------+--------------------|
| |Keeps previous | |
|-m |modification time | |
| |on restored files. | |
|----------------------+-------------------+--------------------|
| * -d |Makes needed | |
| * --make-directories|directories. | |
|----------------------+-------------------+--------------------|
| * -u |Over writes old | |
| * --unconditional |files. | |
|----------------------+-------------------+--------------------|
| | |This option can only|
| | |be used by the root |
| | |user. |
| * -R user group |Reassigns file | |
| * --owner user group|ownership. |If no user is |
| | |specified, the root |
| | |user becomes the |
| | |owner. |
|----------------------+-------------------+--------------------|
|-r |Renames files. |Provide new file |
| | |names when prompted.|
|----------------------+-------------------+--------------------|
| * -t |Displays files in |Use with the -v |
| * --list |the archive. |option to display |
| | |file attributes |
|----------------------+-------------------+--------------------|
| * -v |Displays progress | |
| * --verbose |on-screen. | |
+---------------------------------------------------------------+
The -p option for cpio is useful when you do not want to copy files immediately to the backup device. You might want to prepare the backup first in order to see its size, or to compress the files before backup.
Similarly, if you are creating a cd , copying the files directly to the cd is unreliable or difficult with some cd burning software.
To do a copy for later backup, use cpio with a search command, such as find . See Searching for Files: Using the find Command, Section 8.4.1 .
The basic pattern is:
find -name string -print | cpio -p options > target .
In this example:
The -name option for find lets you search for a string enclosed in double quotation marks. Metacharacters can be used. See Using Metacharacters or Wild Cards, Section 5.3.6.
The bar character ( | ) redirects the output of find to cpio .
The -o flag sets cpio to create an archive file for backing up.
The target is a directory.
The > redirection operator redirects files to the location for the back up. Typically, this location is on a removable device.
The -o Flag in the cpio Command (Backup)
+---------------------------------------------------------+
| Option | Action |Comments|
|-------------+----------------------------------+--------|
| * -A |Adds files to an existing archive.| |
| * --append | | |
|-------------+----------------------------------+--------|
| * -v |Displays progress on-screen. | |
| * --verbose| | |
+---------------------------------------------------------+
The -p Flag for the cpio Command (Delayed Backup)
+---------------------------------------------------------------+
| Option | Action | Comment |
|-----------------------+------------------+--------------------|
| * -a |Resets | |
| * --reset-access-time|modificationtimes.| |
|-----------------------+------------------+--------------------|
| * -d |Makes directories | |
| * --make-directories |as needed. | |
|-----------------------+------------------+--------------------|
| * -f |Keeps previous | |
| * --non-matching |modification time | |
| |on restored files.| |
|-----------------------+------------------+--------------------|
| |Keeps original | |
|-m |modification | |
| |times. | |
|-----------------------+------------------+--------------------|
| | |Use as root user. |
| |Reassigns | |
| * -R user group |ownership of the |If no user is |
| * -- owner user group|file. |specified, then the |
| | |owner becomes the |
| | |root user. |
|-----------------------+------------------+--------------------|
| * -u |Overwrites old | |
| * --unconditional |files. | |
|-----------------------+------------------+--------------------|
| * -v |Displays progress | |
| * --verbose |on-screen. | |
+---------------------------------------------------------------+
You can schedule a command or a script using two tools:
crontab : Schedules tasks once or repeatedly.
at : Schedules tasks once.
You can use the crontab command to run commands at regular times. For example, you could schedule a backup of your files every Friday. Commands can be scheduled to the minute.
[Note]
If you want to schedule a command to run only once, look into the at command.
The crontab command works by reading a crontab file. A crontab file is a plain text file that lists all the jobs you have scheduled.
Each line of the crontab file consists of the command with all its options and arguments, and the date and time to run it.
By default, all users can set up their own crontab files. However, the root user can use two files to control access:
/etc/cron.allow: If this file exists, only users listed in it can use crontab.
/etc/cron.deny: If this file exists, users listed in it cannot use crontab.
All crontab files are stored in the /var/spool/cron/crontabs/ directory. However, since regular users do not have read or write permission to this directory, they must use the crontab command to schedule jobs.
In addition, many system administrators use three files to schedule jobs: etc/cron.daily, /etc/cron.weekly, and /etc/cron.monthly. As the names suggest, these directories contain regularly scheduled jobs that run at daily, weekly or monthly intervals.
To use a crontab file:
1 Type: crontab options user or crontab -e . Only the root user can specify another user's crontab file. The crontab file opens in the vim editor. The first time that you use a file, it is blank.
2 Enter the file, and save it. See Editing and Reading a crontab File, Section 12.6.1.1.
The crontab Command (Scheduling)
+---------------------------------------------------------------+
|Option| Action | Comment |
|------+-----------------------+--------------------------------|
| |Opens a user's crontab |If no user is specified, the |
|-e |file in the vi text |crontab file opened is the |
| |editor. |current user's. |
|------+-----------------------+--------------------------------|
|file |Creates a crontab file.|Mostly useful for placing the |
| | |file in a non-standard location.|
|------+-----------------------+--------------------------------|
|-l |Views a user's crontab |Usable only by the root user. |
|user |file. | |
|------+-----------------------+--------------------------------|
|-r |Removes a user's |Usable only by the root user. |
|user |crontab file. | |
|------+-----------------------+--------------------------------|
|-u |Sets which user's | |
|user |crontab file is |Usable only by the root user. |
| |affected. | |
+---------------------------------------------------------------+
To edit a crontab, type crontab -e user . The file opens in the default editor for the current user.
Each line in a crontab file schedules a single task.
[Note]
If you need to use a percentage sign, put a back slash in front of it: % . Otherwise, it is read as the start of a new line, and the command does not run.
Each line consists of six columns, each one separated by a line or a tab.
Each of the first five columns is an option for setting the date and time. Each of the first five columns can specify:
A particular date or time.
A series of dates or times, with a comma between them.
A range of dates or time, beginning with the start, with a hyphen between them.
See Column Entries in a Crontab File, Section 12.6.1.2.
The sixth column is the command to run, including options and arguments. No quotation marks are need to enclose the command.
For example, 0 12 * * * /usr/share/bin/alarm would run a script called alarm every day at noon.
[Note]
Jobs only run if the time and date match all column settings. Setting column 3 to 1 and column 5 to 1 would only run a job when the first of the month falls on a Monday.
[Note]
No validity check is made. Nothing stops yo, for example, from setting column 3 to 31 and column 4 to 2, setting the time to February 31.
+-------------------------------------------------------------------+
| Column or Option |Argument | Comments |
| | | |
| (from left to | | |
| right) | | |
|------------------+---------+--------------------------------------|
| 1. 1 Minute |0-59 | |
|------------------+---------+--------------------------------------|
| 1. 2 Hour |0-23 |0 is midnight. |
|------------------+---------+--------------------------------------|
| 1. 3 Day of the |1-31 | |
| month | | |
|------------------+---------+--------------------------------------|
| 1. 4 Month |1-12 | |
|------------------+---------+--------------------------------------|
| 1. 5 Day of the |0-6 |0 is Sunday. |
| week | | |
|------------------+---------+--------------------------------------|
| | |Options or arguments can be included. |
| 1. 6 Job |any | |
| |command |No quotation marks are needed around |
| | |the string. |
|------------------+---------+--------------------------------------|
| | |All possible entries. |
|Any column |* | |
| | |Use in the first five columns. |
|------------------+---------+--------------------------------------|
| | |No entry. |
|Any column |- | |
| | |Use in the first five columns. |
+-------------------------------------------------------------------+
You can use the at command to schedule a command or script to run a single time. The command includes several utilities. See Using the at Utilities, Section 12.6.2.1.
[Note]
If you want to schedule a command more than once, look into the crontab command.
By default, only the root user can use the at command. However, the root user can use two files to allow regular users the command:
/etc/at.allow: If this file exists, only users listed in it can use the command.
/etc/at.deny: If this file exists, users listed in it cannot use the command.
The at command depends on a plain text file. This file lists commands or scripts, one per line. The end of the file must be marked by <EOT> on a line by itself.
The time for a task can be set using several formats. The simplest is a 24 hour clock in the hh:mm format, followed by the date in the mm: dd:yyyy format.
You can enter commands by entering:
The name of a plain text file in which commands or scripts are listed, one per line. Type: at -f file command time date .
By entering the time and entering the commands at an at prompt, one per line. The end of the file must be marked by <EOT> on a line by itself. Type: at time date . This method is especially useful if you are entering multiple commands.
The at Command (Scheduling): Options
+---------------------------------------------------------------+
|Option | Action |Comment|
|-------+-----------------------------------------------+-------|
|-c task|Displays listed tasks. | |
|-------+-----------------------------------------------+-------|
|-d task|Deletes the specified tasks. | |
|-------+-----------------------------------------------+-------|
|-f file|Read tasks from the file entered. | |
|-------+-----------------------------------------------+-------|
|-l |Lists all tasks scheduled for the current user.| |
|-------+-----------------------------------------------+-------|
|-m |Sends mail to the user when the task is run. | |
+---------------------------------------------------------------+
The at Command (Scheduling): Time Formats
+---------------------------------------------------------------+
| Option | Action | Comment |
|--------+---------------------------+--------------------------|
| | |The day can be spelled out|
|day |Sets the day of the week. |or abbreviated to its |
| | |first three letters. |
|--------+---------------------------+--------------------------|
|hh:mm | |Use either a 24 hour clock|
|modifier|Sets the time. |or a 12 hour clock, |
| | |modified by am or pm . |
|--------+---------------------------+--------------------------|
|midnight|Sets the time to 12:00 am .| |
|--------+---------------------------+--------------------------|
| | |If the month is before the|
|mmddyyyy|Sets the date. |current month, the command|
| | |runs next year. |
|--------+---------------------------+--------------------------|
|noon |Sets the time to 12:00 pm .| |
|--------+---------------------------+--------------------------|
| |Runs the command | |
|now |immediately or at the same | |
| |time on the specified day. | |
|--------+---------------------------+--------------------------|
|teatime |Sets the time to 4:00 pm . | |
|--------+---------------------------+--------------------------|
|today |Runs the command today. | |
|--------+---------------------------+--------------------------|
|tomorrow|Runs the command tomorrow. | |
+---------------------------------------------------------------+
The at command can be controlled by two utilities:
atrm : Deletes tasks. Type: atrm job .
atq : Lists tasks for the current user, or all jobs if used by the root user. Type: atq .
By default, Debian installs the gui tar program in the main menu. Despite the name, guitar uses the tar, gzip, and gzip2 utilities, as well as several others.
1 From the main menu, select Programs > Utilities > gui tar . The gui tar window opens.
2 Select a task from the tool bar:
New: Create a new archive. You can select:
The tool to use.
The level of compression.
The files to add to the archive.
Open: Open an existing archive.
Close: Close the display of an existing archive.
Add: Add files to an existing archive. You can select the archive from a file manager.
View: Browse an existing archive.
Extract: Choose an archive, and the files to extract from it.
Delete: Choose an archive, and the files to delete from it.
Compressing and archiving files are useful for making backups and e-mail attachments.
Compression reduces the size of files and places them in a single file. Archving creates a single file in which to store multiple files.
In both cases, the individual files cannot be accessed by file managers or normal commands for manipulating files. Instead, you must use special tools designed for the utility that you used for compression.
The processes of decompressing and unarchiving are both referred to as extracting. You can use special commands for extracting, or options in the same tools that you use for compressing or archiving.
Archived files are called tar files, tarred files, or tarballs.
Linux includes several tools for compressing and archiving:
tar : mainly an archive tool, but also compresses files. Many long time users prefer tar to the alternatives.
gzip : a compression tool. Sometimes, gzip can decompress a zip file from another operating system, but do not confuse the two formats or expect them to be always interchangeable.
bzip2 : another compression tool, The bzip2 tool is capable of greater compression than gzip , and includes a recovery tool, but less widely used.
In addition, Linux includes tools for viewing, searching, and comparing files compressed using gzip .
The tar command can compress as well as archive files. Its -z option is equivalent to using the gzip command.
An archive file has a .tar extension. An archive file that is also compressed has a .tgz extension. At times, an archived and compressed file will have a tar.gz extension
The tar command is some times used for back ups. However, cpio is often used instead, because it offers more control over back ups. See Checking or Repairing a File System, Section 12.9.
Type: tar options archive files .
The archive is the tar file to create or extract from. If it does not exist, it is created by the command.
Separate each file name by a space.
The tar Command (Archive)
+-------------------------------------------------------------------+
| Option | Action | Comment |
|--------------------------+----------------+-----------------------|
| |Keep original | |
| * --atime-preserve |time stamps on | |
| |extracted files.| |
|--------------------------+----------------+-----------------------|
| * -c |Creates a new | |
| * --create |archive file. | |
|--------------------------+----------------+-----------------------|
| * -d tarfile files |Compare the |Missing files, sizes |
| * --compares tarfile |files in a tar |and attributes are |
| files |file with other |reported on. |
| |files. | |
|--------------------------+----------------+-----------------------|
| |Excludes a file | |
| * --exclude file |from archiving | |
| |or extracting. | |
|--------------------------+----------------+-----------------------|
| |Extracts files | |
| * -k |without |An error message |
| * --keep-old-files |overwriting |displays if there are |
| |files with the |conflicts. |
| |same name. | |
|--------------------------+----------------+-----------------------|
| |Splits over | |
| * -M |multiple |Works with tapes or |
| * --multivolume |removable |floppies. |
| |devices. | |
|--------------------------+----------------+-----------------------|
| |Excludes files | |
| * -N date |older than the | |
| * --after-date date |date entered | |
| |from the | |
| |archive. | |
|--------------------------+----------------+-----------------------|
| * -P |Keeps absolute |See Understanding |
| * --absolute-paths |paths. |Absolute and Relative |
| | |File Paths. (***1) |
|--------------------------+----------------+-----------------------|
| |Keeps | |
| * -p |permissions of |See Changing |
| * --same-permissions |files unchanged |Permissions. (***2) |
| * --preserve-permissions|when they are | |
| |archived. | |
|--------------------------+----------------+-----------------------|
| |Adds a tar file |The second file listed |
| * -r |to the end of |in the command is added|
| * --append |another one. |to the end of the |
| | |first. |
|--------------------------+----------------+-----------------------|
| |Removes files | |
| * --remove-files |after they are | |
| |added to a tar | |
| |file. | |
|--------------------------+----------------+-----------------------|
| * -t |Displays files | |
| * --list |in an archive. | |
|--------------------------+----------------+-----------------------|
| | |Files are only added if|
| * -u |Updates an |they are new than files|
| * --update |archive. |of the same name that |
| | |are already in the |
| | |archive. |
|--------------------------+----------------+-----------------------|
| * -v |Displays actions| |
| * --verbose |on-screen. | |
|--------------------------+----------------+-----------------------|
| |Asks | |
| * -w |confirmation |Enter y for Yes, and n |
| * --interactive |before working |for No. |
| |with each file. | |
|--------------------------+----------------+-----------------------|
| |Checks for | |
| * -W |errors in | |
| * --verify |archiving or | |
| |extracting. | |
|--------------------------+----------------+-----------------------|
| * -x |Extracts files | |
| * --extract |from an archive.| |
| * --get | | |
|--------------------------+----------------+-----------------------|
| * -z |Compress or |--compress can also be |
| * -Z |uncopresses |used to compress. |
| |using gzip. | |
+-------------------------------------------------------------------+
***1) See Understanding Absolute and Relative File Paths, Section 8.2.2. ***2) See Changing Permissions, Section 9.19.
The gzip command is one of the two Linux tools for compressing and decompressing files. The other is bzip2 . The bzip2 command can compress files more, but gzip is more common, and can sometimes read zipped files, or compressed files created with the obsolete tool compress .
A file compressed using gzip has a .gz extension.
Type: gzip options files .
Separate file names with a space.
The gzip command has several related utilities:
zcat : Displays files in a gzip archive.
zcmp : Compares files in a gzip archive.
zdiff : Compares files in a gzip archive.
zgrep, zegrep, zfgrep : Search for text patterns in files in a gzip archive.
gunzip : decompresses files. However, gzip -d also decompresses.
All these commands have the same format as gzip . They can use the same options as the versions of the commands used for uncompressed files.
The gzip Command (Compress)
+-------------------------------------------------------------------+
| Option | Action | Comment |
|----------------+------------------------+-------------------------|
| |Sets the amount of |The default setting is 6.|
| |compression to greatest | |
|--best |setting (on a scale of 1|When the compression is |
| |to 9). |greatest, the speed of |
| | |compression is slowest. |
|----------------+------------------------+-------------------------|
| * -d |Extracts compressed | |
| * --decompress|files. | |
| * --decompress| | |
|----------------+------------------------+-------------------------|
| |Sets the compression to |When the speed of |
|--fastest |the fastest setting (1 |compression is greatest, |
| |on a scale of 1 to 9). |the amount of compression|
| | |is least. |
|----------------+------------------------+-------------------------|
| * -f |Forces compression. | |
| * --force | | |
|----------------+------------------------+-------------------------|
| * -N |Keeps original name and | |
| * --name |time stamp. | |
|----------------+------------------------+-------------------------|
| * -n |Discards original name |This is the default |
| * --no-name |and time stamp. |setting for |
| | |decompressing. |
|----------------+------------------------+-------------------------|
| |Sets the speed and |1 is the fastest, 9 the |
|-number |amount of compression on|most compressed. |
| |a scale of 1 t0 9. | |
| | |The default setting is 6.|
|----------------+------------------------+-------------------------|
| * -r |Includes sub-directories| |
| * --recursive |and their contents. | |
|----------------+------------------------+-------------------------|
| * -v |Displays progress |The file and the amount |
| * --verbose |on-screen. |of compression display. |
+-------------------------------------------------------------------+
The bzip2 command is one of two tools for compressing and decompressing files. The other is gzip . The gzip command is more common, but bzip2 's compression is 10 to 20 percent greater, depending on the file.
Compressed files made with bzip2 have a .bz2 extension.
Type: bzip2 options files .
Separate each name with a space.
The bzip2 command has three related utilities:
bzipcat : Displays the contents of a file in a bzip compressed file. Uses the same options as the cat command.
bziprecover : Helps to recover damaged compressed files.
bunzip2 : decompresses files. However, many users prefer bzip2 -d for decompressing files.
All these utilities have the same command format as bzip2 .
The bzip2 Command (Compress)
+-------------------------------------------------------------------+
| Option | Action |Comments|
|----------------+-----------------------------------------+--------|
| * -d |Decompresses files. | |
| * --decompress| | |
|----------------+-----------------------------------------+--------|
| * -f |Over writes existing files. | |
| * --force | | |
|----------------+-----------------------------------------+--------|
| * -k |Keeps original files once a compressed | |
| * --keep |file is made. | |
|----------------+-----------------------------------------+--------|
| * -v |Displays progress on screen. | |
| * --verbose | | |
|----------------+-----------------------------------------+--------|
| * -z |Compresses files. | |
| * --compress | | |
+-------------------------------------------------------------------+
The native Linux filesystem (ext2) does not need to be defragmented. However, occasionally, you may need to check a partition's file allocation and make repairs.
Type: fsck options filesystem or fsck options mountpoint .
Note that:
By default, repairs are done without prompting.
Two ways of specifying a filesystem exist. You do not need to specify a filesystem if you use the -t or -A options.
This check is also run when the system starts if it was not shut down gracefully or after a predefined number or reboots.
[Note]
Linux partitions do not need defragmenting. fat 16 and fat 32 filesystems can be defragmented under another operating system.
The fsck Command (Checks Filesystems)
+-------------------------------------------------------------------+
|Option| Action | Comment |
|------+----------------------------+-------------------------------|
|-A |Checks all filesystems | |
| |listed in /etc/fstab. | |
|------+----------------------------+-------------------------------|
|-r |Prompts before each repairs | |
| |is done. | |
|------+----------------------------+-------------------------------|
| |Checks all filesystems | |
|-R |listed in /etc/fstab except |Use with -A. |
| |the root partition. | |
|------+----------------------------+-------------------------------|
| | |Common types are: |
| | | |
| | | * ext2 : The native Linux |
| | | format. |
|-t |To only check filesystems of| * swap : Linux swap file. Use|
|type |a certain type | only for the swap file. |
| | | * vfat :The format for |
| | | Windows 98, me , and the |
| | | second release of Windows |
| | | 95. |
| | | * iso9660: cd drive |
+-------------------------------------------------------------------+
[ previous ] [ Contents ] [ 1 ] [ 2 ] [ 3 ] [ 4 ] [ 5 ] [ 6 ] [ 7 ] [ 8 ] [ 9 ] [ 10 ] [ 11 ] [ 12 ] [ 13 ] [ A ] [ B ] [ C ] [ D ] [ next ]
User's Guide (Obsolete Documentation) (Obsolete Documentation)
Version: 1.00p00, 2009.07.21-11:14