Process Management

The operating system (OS) manages multiple user requests and tasks. In most cases, the OS comes with only one CPU and one main memory, but multiple tier-2 disks and input/output (I/O) devices. Therefore, users have to share resources, but it appears to users that they are exclusively occupying resources. The OS places user tasks, OS tasks, mailing, print tasks, and other pending tasks in a queue and schedules the tasks according to predefined rules. This topic describes how the OS manages processes.

Viewing Processes

Linux is a multi-task system and needs to get process information during process management. To manage processes, you need to know the number of processes and their statuses. Multiple commands are available to view processes.

who Command

The who command is used to display system user information. For example, before running the talk command to establish instant communication with another user, you need to run the who command to determine whether the target user is online. In another example, the system administrator can run the who command to learn what each login user is doing at the current time. The who command is widely seen in system administration since it is easy to use and can return a comprehensive set of accurate user information.

The following is an example output of the who command, where system users and their status are displayed: The use of the who command is as follows:

$ who
admin     tty1         Jul 28 15:55
admin     pts/0        Aug  5 15:46 (192.168.0.110)
admin     pts/2        Jul 29 19:52 (192.168.0.110)
root     pts/3        Jul 30 12:07 (192.168.0.110)
root     pts/4        Jul 31 10:29 (192.168.0.144)
root     pts/5        Jul 31 14:52 (192.168.0.11)
root     pts/6        Aug  6 10:12 (192.168.0.234)
root     pts/8        Aug  6 11:34 (192.168.0.234)

ps Command

The ps command is the most basic and powerful command to view process information, including which processes are running, terminated, resource-hungry, or stay as zombies.

A common scenario is to monitor background processes, which do not interact with your screen, keyboard, and other I/O devices. Table 1 lists the common ps command options.

Table 1 Common ps command options

Option

Description

-e

Displays all processes.

-f

Full output format.

-h

Hides column headings in the process information.

-l

Long output format.

-w

Wide output format.

-a

Lists all processes on a terminal, including those of other users.

-r

Lists only running processes.

-x

Lists all processes without control terminals.

For example, to list all processes on a terminal, run the following command:

$ ps -a
  PID TTY          TIME CMD
12175 pts/6    00:00:00 bash
24526 pts/0    00:00:00 vsftpd
29478 pts/5    00:00:00 ps
32461 pts/0    1-01:58:33 sh

top Command

Both the top and ps commands can display a list of currently running processes, but the top command allows you to update the displayed list of processes by pressing a button repeatedly. If the top command is executed in foreground, it exclusively occupies foreground until it is terminated. The top command provides real-time visibility into system processor status. You can sort the list of CPU tasks by CPU usage, memory usage, or task execution time. Extensive display customization, such as choosing the columns or sorting method, can be achieved using interactive commands or the customization file.

Figure 1 provides an example output of the top command.

Figure 1 Example command output

kill Command

The kill command is used to terminate a process regardless of whether the process is running in foreground or background. It differs from the combo key Ctrl+C, which can terminate only foreground processes. The reason for terminating a background process can be heavy use of CPU resources or deadlock.

The kill command sends a signal to terminate running processes. By default, the TERM signal is used, terminating all processes incapable of capturing it. To terminate a process capable of capturing the TERM signal, use the KILL signal (signal ID: 9) instead.

Two types of syntax of the kill command:

kill [-s signal | -p] [-a] PID…
kill -l [signal]

The process ID can be retrieved by running the ps command. The -s option indicates the signal sent to the specified program. The signal details can be viewed by running the kill -l command. The -p option indicates the specified process ID.

For example, to terminate the process whose ID is 1409, run the following command as the root user:

# kill -9 1409

Example output of the kill command with the -l option

$ kill -l
 1) SIGHUP       2) SIGINT       3) SIGQUIT      4) SIGILL       5) SIGTRAP
 6) SIGABRT      7) SIGBUS       8) SIGFPE       9) SIGKILL     10) SIGUSR1
11) SIGSEGV     12) SIGUSR2     13) SIGPIPE     14) SIGALRM     15) SIGTERM
16) SIGSTKFLT   17) SIGCHLD     18) SIGCONT     19) SIGSTOP     20) SIGTSTP
21) SIGTTIN     22) SIGTTOU     23) SIGURG      24) SIGXCPU     25) SIGXFSZ
26) SIGVTALRM   27) SIGPROF     28) SIGWINCH    29) SIGIO       30) SIGPWR
31) SIGSYS      34) SIGRTMIN    35) SIGRTMIN+1  36) SIGRTMIN+2  37) SIGRTMIN+3
38) SIGRTMIN+4  39) SIGRTMIN+5  40) SIGRTMIN+6  41) SIGRTMIN+7  42) SIGRTMIN+8
43) SIGRTMIN+9  44) SIGRTMIN+10 45) SIGRTMIN+11 46) SIGRTMIN+12 47) SIGRTMIN+13
48) SIGRTMIN+14 49) SIGRTMIN+15 50) SIGRTMAX-14 51) SIGRTMAX-13 52) SIGRTMAX-12
53) SIGRTMAX-11 54) SIGRTMAX-10 55) SIGRTMAX-9  56) SIGRTMAX-8  57) SIGRTMAX-7
58) SIGRTMAX-6  59) SIGRTMAX-5  60) SIGRTMAX-4  61) SIGRTMAX-3  62) SIGRTMAX-2
63) SIGRTMAX-1  64) SIGRTMAX

Scheduling a Process

The time-consuming and resource-demanding part of maintenance work is often performed at late night. You can schedule relevant processes to get started at the scheduled time instead of staying up all night. The following describes the process scheduling commands.

Using the at Command to Run Processes at the Scheduled Time

Function

The at command is used to run a batch of processes (a series of commands) at the scheduled time or time and date.

Syntax of the at command:

at [-V] [-q queue] [-f filename] [-mldbv] time
at -c job [job...]

Time Format

The scheduled time can be in any of the following formats:

  • hh:mm today: If hh:mm is earlier than the current time, the selected commands will be run at hh:mm the next day.
  • midnight, noon, teatime (typically at 16:00), or the like
  • 12-hour format followed by am or pm
  • Time + date (month day, mm/dd/yy, or dd.mm.yy). The scheduled date must follow the scheduled time.

The scheduled time can also be relative time, which is suitable for scheduling commands that are going to be executed soon. For example, now+N minutes, hours, days, or weeks. N indicates the specified time, which may be a few days or hours. Further, the scheduled time can be words like today, tomorrow, or the like. Here are some examples of the scheduled time.

Assume that the current time is 12:30 June 7 2019 and you want to run a command at 4:30 pm. The time scheduled by the at command can be any of the following:

 at 4:30pm
 at 16:30
 at 16:30 today
 at now+4 hours
 at now+ 240 minutes
 at 16:30 7.6.19
 at 16:30 6/7/19
 at 16:30 Jun 7

Although you can select any of the preceding examples according to your preference, absolute time in 24-hour format, such as at 16:30 6/7/19, is recommended.

Privileges

Only commands from standard input or from the file specified by the -f option can be scheduled by the at command. If the su command is executed to switch the OS from user A to user B and then the at command is executed at the shell prompt of user B, the at command execution result is sent to user B, whereas emails (if any) are sent to user A.

For example, to run the slocate -u command at 10 am on June 8, 2019, run the following commands as the root user:

# at  10:00  6/8/19
at> slocate -u
at>
[1]+   Stopped    at  10:00  6/8/19

When the at> prompt appears, type slocate -u and press Enter. Repeat the step to add other commands that need to be run at 10 am on 8 June 2019. Then, press Ctrl+D to exit the at command.

The administrator is authorized to run the at command unconditionally. For other users, their privileges to run the at command is defined in the /etc/at.allow and /etc/at.deny files.

Using the cron Service to Run Commands Periodically

The at command can run commands at the scheduled time, but only once. It means that after the commands to be run is specified, the system completes the task at the specified time. If you need to run the commands repeatedly, the cron service is a good choice.

Cron Service

The cron service searches the /var/spool/cron directory for the crontab files named by the user name in the /etc/passwd file and loads the search results into memory to execute the commands in the crontab files. Each user has a crontab file with the same name as the user name. For example, the crontab file of the userexample user is /var/spool/cron/userexample.

The cron service also reads the cron configuration file /etc/crontab every minute, which can be edited in various formats. If no crontab files are found, the cron service enters sleep mode and releases system resources. One minute later, the cron service is waken up to repeat the search work and command execution. Therefore, the background process occupies few resources and is wakened up every minute to check whether there are commands to be executed.

Command execution results are then mailed to users specified by the environment variable MAILTO in the /etc/crontab file. The cron service, once started, does not require manual intervention except when you need to replace the scheduled commands with new ones.

crontab Command

The crontab command is used to install, edit, remove, list, and perform other operations on crontab files. Each user has its own crontab files and can add commands to be executed to the files.

Here are common crontab command options:

  • crontab -u //Set the cron service of a user. This option is required only when the crontab command is run by the root user.
  • crontab -l //List details about the cron service of a user.
  • crontab -r //Remove the cron service of a user.
  • crontab -e //Edit the cron service of a user.

For example, to list the cron service settings of the root user, run the following command:

# crontab -u root -l

crontab Files

Enter the commands to be executed and their scheduled time in crontab files. Each line in the files contains six fields. The first five fields are the time when the specified command is executed, and the last field is the command to be executed. Fields are separated by spaces or tabs. The format is as follows:

minute hour day-of-month month-of-year day-of-week commands

The following table describes the fields in each line.

Table 2 Parameter description

Parameter

Description

minute

The minute of the hour at which commands will be executed. Value range: 0–59.

hour

The hour of the day at which scheduled commands will be executed. Value range: 0–23.

day-of-month

The day of the month on which scheduled commands will be executed. Value range: 1–31.

month-of-year

The month of the year in which scheduled commands will be executed. Value range: 1–12.

day-of-week

The day of the week on which scheduled commands will be executed. Value range: 0–6.

commands

Scheduled commands.

The fields cannot be left unspecified. In addition to numerical values, the following special characters are allowed: asterisk (*), indicating a wildcard value; forward slash (/), followed by a numeral value N to indicate that commands will be executed at a regular interval of N; hyphen (-), used with a range; and comma (,), used to separate discrete values. A complete path to the commands must be provided.

For example, to allow the OS to add sleepy to the /tmp/test.txt file every two hours from 18 pm to 22 pm, add the following line to a crontab file:

* 18-22/2 * * * echo "sleepy" >> /tmp/test.txt

Each time the cron service settings of a user are edited, the cron service generates a crontab file with the same name as the user in the /var/spool/cron directory. The crontab file can be edited only using the crontab -e command. Alternatively, the user can create a file and run the crontab _filename_ command to import its cron settings to the new file.

For example, to create a crontab file for the userexample user, perform the following steps:

  1. Create a file using any text editor. Add the commands that need to be executed periodically and the command execution interval to the new file. In this example, the new file is ~/userexample.cron.

  2. Run the following command as the root user to install the new file as the crontab file of the userexample user:

    # crontab -u userexample ~/userexample.cron
    

After the new file is installed, you will find a file named userexample in the /var/spool/cron directory. This file is the required crontab file.

NOTE:
Do not restart the cron service after a crontab file is modified, because the cron service, once started, reads the crontab file every minute to check whether there are commands that need to be executed periodically.

/etc/crontab File

The cron service reads all files in the /var/spool/cron directory and the /etc/crontab file every minute. Therefore, you can use the cron service by configuring the /etc/crontab file. A crontab file contains user-specific commands, whereas the /etc/crontab file contains system-wide commands. The following is an example of the /etc/crontab file.

SHELL=/bin/sh
PATH=/usr/bin:/usr/sbin:/sbin:/bin:/usr/lib/news/bin
MAILTO=root //If an error occurs or data is output, the data is sent to the account by email.
HOME=/
#  run-parts
01 * * * * root run-parts /etc/cron.hourly //Run scripts in the /etc/cron.hourly directory once an hour.
02 4 * * *   root run-parts /etc/cron.daily    //Run scripts in the /etc/cron.daily directory once a day.
22 4 * * 0  root run-parts /etc/cron.weekly     //Run scripts in the /etc/cron.weekly directory once a week.
42 4 1  * *  root run-parts /etc/cron.monthly     //Run scripts in the /etc/cron.monthly directory once a month.

NOTE:
If the run-parts parameter is deleted, a script name instead of a directory name is used.

Suspending/Resuming a Process

A process can be suspended or resumed by job control, and the process will continue to work from the suspended point after being resumed. To suspend a foreground process, press Ctrl+Z. After you press Ctrl+Z, the cat command is suspended together with the foreground process you want to suspend. You can use the jobs command instead to display a list of shell jobs, including their names, IDs, and status.

To resume a process in foreground or background, run the fg or bg command, respectively. The process then starts from where it was suspended previously.

有奖捉虫

“有虫”文档片段

存在的问题

提交类型 issue
有点复杂...
找人问问吧。
PR
小问题,全程线上修改...
一键搞定!
问题类型
规范和低错类

● 错别字或拼写错误;标点符号使用错误;

● 链接错误、空单元格、格式错误;

● 英文中包含中文字符;

● 界面和描述不一致,但不影响操作;

● 表述不通顺,但不影响理解;

● 版本号不匹配:如软件包名称、界面版本号;

易用性

● 关键步骤错误或缺失,无法指导用户完成任务;

● 缺少必要的前提条件、注意事项等;

● 图形、表格、文字等晦涩难懂;

● 逻辑不清晰,该分类、分项、分步骤的没有给出;

正确性

● 技术原理、功能、规格等描述和软件不一致,存在错误;

● 原理图、架构图等存在错误;

● 命令、命令参数等错误;

● 代码片段错误;

● 命令无法完成对应功能;

● 界面错误,无法指导操作;

风险提示

● 对重要数据或系统存在风险的操作,缺少安全提示;

内容合规

● 违反法律法规,涉及政治、领土主权等敏感词;

● 内容侵权;

您对文档的总体满意度

非常不满意
非常满意
创Issue赢奖品
根据您的反馈,会自动生成issue模板。您只需点击按钮,创建issue即可。