1 00:00:00,270 --> 00:00:01,350 In this lesson, 2 00:00:01,350 --> 00:00:04,046 we're going to discuss how you can troubleshoot process issues 3 00:00:04,046 --> 00:00:06,600 within the Linux operating system. 4 00:00:06,600 --> 00:00:09,480 While running processes within the Linux operating system, 5 00:00:09,480 --> 00:00:12,330 your users may experience technical issues. 6 00:00:12,330 --> 00:00:14,970 To provide uninterrupted services to your users, 7 00:00:14,970 --> 00:00:16,703 you may need to be able to solve problems 8 00:00:16,703 --> 00:00:19,320 that are going to come up during operations. 9 00:00:19,320 --> 00:00:22,075 As an administrator managing multiple systems on a network, 10 00:00:22,075 --> 00:00:23,940 you're going to have installed and managed 11 00:00:23,940 --> 00:00:25,770 various services and packages 12 00:00:25,770 --> 00:00:28,110 that are required for your users to do their job. 13 00:00:28,110 --> 00:00:30,630 However, there's sometimes going to be instances 14 00:00:30,630 --> 00:00:32,280 where applications and services 15 00:00:32,280 --> 00:00:34,860 do not function as they're supposed to. 16 00:00:34,860 --> 00:00:36,390 As a system administrator, 17 00:00:36,390 --> 00:00:38,160 you're going to be expected to determine 18 00:00:38,160 --> 00:00:40,110 and resolve those problems. 19 00:00:40,110 --> 00:00:42,300 Now there are many different issues that could affect 20 00:00:42,300 --> 00:00:45,630 or be caused by a process of running instance of software 21 00:00:45,630 --> 00:00:47,340 or something else like that. 22 00:00:47,340 --> 00:00:50,790 These issues can cause instability when processes hang. 23 00:00:50,790 --> 00:00:52,200 They can consume resources 24 00:00:52,200 --> 00:00:54,180 that should be allocated to other processes 25 00:00:54,180 --> 00:00:57,060 and it can create a general sluggishness to your system 26 00:00:57,060 --> 00:00:58,890 and other issues like that. 27 00:00:58,890 --> 00:01:00,720 Processes move through a life cycle 28 00:01:00,720 --> 00:01:03,630 from creation until their eventual termination. 29 00:01:03,630 --> 00:01:06,180 Now there are five different states that a process can be in 30 00:01:06,180 --> 00:01:07,710 during this life cycle. 31 00:01:07,710 --> 00:01:10,860 Each one defines how the system or other applications 32 00:01:10,860 --> 00:01:13,230 can interact with that given process. 33 00:01:13,230 --> 00:01:14,820 Knowing a process's state 34 00:01:14,820 --> 00:01:17,130 can help you determine what to do with that process 35 00:01:17,130 --> 00:01:19,350 or it can help you diagnose problems 36 00:01:19,350 --> 00:01:21,210 concerning that process. 37 00:01:21,210 --> 00:01:24,180 The running state means the process is currently executing 38 00:01:24,180 --> 00:01:27,030 in the user space or the kernel space. 39 00:01:27,030 --> 00:01:29,730 The interruptible sleep state means the process 40 00:01:29,730 --> 00:01:31,950 relinquishes access to the CPU 41 00:01:31,950 --> 00:01:34,920 and is waiting to be reactivated by the scheduler. 42 00:01:34,920 --> 00:01:36,660 The interruptible sleep state 43 00:01:36,660 --> 00:01:38,550 means that the process will only wake up 44 00:01:38,550 --> 00:01:40,560 when its resources that it's waiting for 45 00:01:40,560 --> 00:01:42,000 become available to it. 46 00:01:42,000 --> 00:01:45,060 Otherwise, it's going to stay in a sleep state. 47 00:01:45,060 --> 00:01:47,040 Then, we have the zombie state 48 00:01:47,040 --> 00:01:49,058 and this indicates that a process was terminated 49 00:01:49,058 --> 00:01:52,620 but it has not yet been released by its parent process 50 00:01:52,620 --> 00:01:55,020 so it cannot accept a kill signal. 51 00:01:55,020 --> 00:01:56,850 The final state is a stop state 52 00:01:56,850 --> 00:01:58,662 and this state indicates that a process 53 00:01:58,662 --> 00:02:02,220 was stopped by a debugger or through a kill signal. 54 00:02:02,220 --> 00:02:04,170 Now every process on your system 55 00:02:04,170 --> 00:02:07,680 has a unique process ID known as a PID 56 00:02:07,680 --> 00:02:10,169 and it gets this when it started up by the system, 57 00:02:10,169 --> 00:02:11,756 and that way, users and the system 58 00:02:11,756 --> 00:02:14,190 can identify that process. 59 00:02:14,190 --> 00:02:17,370 The PID is a non-negative integer that increases 60 00:02:17,370 --> 00:02:20,490 for each new process that's being started on the system. 61 00:02:20,490 --> 00:02:23,490 Now the init daemon always has a PID of one 62 00:02:23,490 --> 00:02:26,370 because it's the first process to start up on the system 63 00:02:26,370 --> 00:02:28,740 and it becomes the parent for all other processes 64 00:02:28,740 --> 00:02:30,390 on that system. 65 00:02:30,390 --> 00:02:31,770 After this one is started, 66 00:02:31,770 --> 00:02:34,140 then processes are going to be able to be started up 67 00:02:34,140 --> 00:02:36,150 by the system or the user 68 00:02:36,150 --> 00:02:37,560 by that init daemon, 69 00:02:37,560 --> 00:02:39,840 and they're going to be assigned a higher available number 70 00:02:39,840 --> 00:02:42,090 because they're going to be something higher than one 71 00:02:42,090 --> 00:02:44,310 because they're started by the init daemon. 72 00:02:44,310 --> 00:02:45,990 When you're troubleshooting something, 73 00:02:45,990 --> 00:02:48,540 you need to know what the process's PID is 74 00:02:48,540 --> 00:02:50,220 or that identification number 75 00:02:50,220 --> 00:02:52,202 because this helps you understand how to terminate it, 76 00:02:52,202 --> 00:02:55,320 change its priority or perform other management tasks 77 00:02:55,320 --> 00:02:57,330 on that given process. 78 00:02:57,330 --> 00:02:58,586 Now if you want to find that, 79 00:02:58,586 --> 00:03:01,017 you can use the pgrep command. 80 00:03:01,017 --> 00:03:02,550 The pgrep command 81 00:03:02,550 --> 00:03:04,620 looks through the currently running processes 82 00:03:04,620 --> 00:03:05,997 and lists out the process IDs 83 00:03:05,997 --> 00:03:07,898 which match the selection criteria 84 00:03:07,898 --> 00:03:09,930 to your standard output, 85 00:03:09,930 --> 00:03:11,520 similar to how grep uses it 86 00:03:11,520 --> 00:03:14,610 to find matching patterns inside of a text file. 87 00:03:14,610 --> 00:03:16,020 If you use pgrep, 88 00:03:16,020 --> 00:03:17,790 you're going to be able to identify a process 89 00:03:17,790 --> 00:03:19,401 based on multiple different factors 90 00:03:19,401 --> 00:03:21,633 even if you don't know its exact PID. 91 00:03:22,560 --> 00:03:24,750 This will help you identify the PID 92 00:03:24,750 --> 00:03:27,840 and identifying the PID is typically your first step 93 00:03:27,840 --> 00:03:29,520 in managing a process. 94 00:03:29,520 --> 00:03:31,110 To run the pgrep command, 95 00:03:31,110 --> 00:03:33,074 simply type in pgrep, the options, 96 00:03:33,074 --> 00:03:35,094 and the pattern you're searching for. 97 00:03:35,094 --> 00:03:37,830 Another way to find the process idea of a running program 98 00:03:37,830 --> 00:03:39,750 is to use the pidof command, 99 00:03:39,750 --> 00:03:41,934 pidof, spelled P-I-D-O-F, 100 00:03:41,934 --> 00:03:44,340 is going to allow you to find the process ID 101 00:03:44,340 --> 00:03:46,073 of the named program that you give it. 102 00:03:46,073 --> 00:03:48,600 To use it, you're simply going to type pidof, 103 00:03:48,600 --> 00:03:50,010 and then the name of the program. 104 00:03:50,010 --> 00:03:51,900 For example, if you are running something 105 00:03:51,900 --> 00:03:53,400 like the ls command, 106 00:03:53,400 --> 00:03:56,340 you could run pidof space ls, Enter, 107 00:03:56,340 --> 00:03:57,900 and then find what was the PID 108 00:03:57,900 --> 00:04:00,390 associated with the ls command. 109 00:04:00,390 --> 00:04:03,420 Another command you can use is known as ps. 110 00:04:03,420 --> 00:04:06,330 The ps command invokes the process table 111 00:04:06,330 --> 00:04:07,740 which is a record that summarizes 112 00:04:07,740 --> 00:04:10,710 the current running processes on a given system. 113 00:04:10,710 --> 00:04:12,900 When the command is run without any options, 114 00:04:12,900 --> 00:04:14,412 it's going to display all the processes 115 00:04:14,412 --> 00:04:17,519 that are run by the current shell with the details. 116 00:04:17,519 --> 00:04:19,529 This includes things like the PID, 117 00:04:19,529 --> 00:04:21,241 the terminal associated with that process, 118 00:04:21,241 --> 00:04:23,490 the accumulated CPU time, 119 00:04:23,490 --> 00:04:26,190 and the command that started that exact process. 120 00:04:26,190 --> 00:04:28,076 However, different options may be used 121 00:04:28,076 --> 00:04:29,580 along with that command 122 00:04:29,580 --> 00:04:33,000 to filter out the displayed fields or processes too. 123 00:04:33,000 --> 00:04:34,590 The a option, for example, 124 00:04:34,590 --> 00:04:36,730 will list all user-triggered processes. 125 00:04:36,730 --> 00:04:39,360 The -e option is going to be used to list 126 00:04:39,360 --> 00:04:40,830 all of the processes. 127 00:04:40,830 --> 00:04:43,200 The -l option will list the processes 128 00:04:43,200 --> 00:04:45,180 using a long-listing format. 129 00:04:45,180 --> 00:04:47,927 And the u option is going to be used to list out processes 130 00:04:47,927 --> 00:04:50,490 along with the username and the start time. 131 00:04:50,490 --> 00:04:53,610 The r option is going to be used to exclude processes 132 00:04:53,610 --> 00:04:55,200 that are not running currently. 133 00:04:55,200 --> 00:04:57,690 And the x option is used to include processes 134 00:04:57,690 --> 00:04:59,040 without a terminal. 135 00:04:59,040 --> 00:05:01,568 The T option is going to be used to exclude processes 136 00:05:01,568 --> 00:05:03,330 that were started by any terminal 137 00:05:03,330 --> 00:05:04,770 other than the current one. 138 00:05:04,770 --> 00:05:06,976 And the dash -U option with the username 139 00:05:06,976 --> 00:05:09,240 is going to be used to display all the processes 140 00:05:09,240 --> 00:05:11,520 based on a specified user. 141 00:05:11,520 --> 00:05:15,270 The -p and PID option will display only the process 142 00:05:15,270 --> 00:05:17,610 associated with the specific PID. 143 00:05:17,610 --> 00:05:20,769 The -C command option is going to display all the processes 144 00:05:20,769 --> 00:05:22,710 by their command name 145 00:05:22,710 --> 00:05:25,413 while the --tty internal number option 146 00:05:25,413 --> 00:05:27,634 is going to be used to display all the processes 147 00:05:27,634 --> 00:05:30,360 running on a specified terminal. 148 00:05:30,360 --> 00:05:32,820 Like ps, the top command is going to list 149 00:05:32,820 --> 00:05:35,490 all the processes running on a Linux system. 150 00:05:35,490 --> 00:05:37,650 It acts as a process management tool 151 00:05:37,650 --> 00:05:39,840 by enabling you to prioritize, sort, 152 00:05:39,840 --> 00:05:42,450 or terminate processes interactively. 153 00:05:42,450 --> 00:05:44,760 It's going to display a dynamic process status 154 00:05:44,760 --> 00:05:46,860 and reflect real-time changes. 155 00:05:46,860 --> 00:05:49,500 Different keystrokes within this tool are going to execute 156 00:05:49,500 --> 00:05:51,550 various process management actions as well. 157 00:05:51,550 --> 00:05:54,000 Some of the frequently used command keys 158 00:05:54,000 --> 00:05:55,320 include the Enter key. 159 00:05:55,320 --> 00:05:56,190 And when you press that, 160 00:05:56,190 --> 00:05:59,040 it's going to refresh the status of all your processes. 161 00:05:59,040 --> 00:06:00,810 If you press Shift + N, 162 00:06:00,810 --> 00:06:02,220 you're going to sort the processes 163 00:06:02,220 --> 00:06:04,382 in decreasing order of their PID. 164 00:06:04,382 --> 00:06:05,886 If you press the M key, 165 00:06:05,886 --> 00:06:09,210 it's going to sort the processes by their memory usage. 166 00:06:09,210 --> 00:06:10,379 If you press the P key, 167 00:06:10,379 --> 00:06:13,314 it's going to sort the processes by CPU usage. 168 00:06:13,314 --> 00:06:16,620 If you press U, it's going to display the processes 169 00:06:16,620 --> 00:06:19,260 belonging to the user specified at the prompt. 170 00:06:19,260 --> 00:06:20,490 If you press K, 171 00:06:20,490 --> 00:06:22,050 it's going to terminate the process 172 00:06:22,050 --> 00:06:24,420 for which you specify the PID for. 173 00:06:24,420 --> 00:06:25,620 If you press R, 174 00:06:25,620 --> 00:06:27,420 it's going to alter the scheduling priority 175 00:06:27,420 --> 00:06:31,290 of that process and you have to specify this by its PID. 176 00:06:31,290 --> 00:06:32,370 And if you press Q, 177 00:06:32,370 --> 00:06:34,530 it's going to exit the process list. 178 00:06:34,530 --> 00:06:36,946 To run the top command, simply type in top, 179 00:06:36,946 --> 00:06:39,780 and your options at the command prompt. 180 00:06:39,780 --> 00:06:40,980 Now in addition to top, 181 00:06:40,980 --> 00:06:43,440 there's a newer version called htop, 182 00:06:43,440 --> 00:06:46,230 htop is also an interactive, system monitor, 183 00:06:46,230 --> 00:06:48,074 process viewer and process manager 184 00:06:48,074 --> 00:06:50,280 but it really does a lot more 185 00:06:50,280 --> 00:06:53,219 and in a lot easier way of using it than does top. 186 00:06:53,219 --> 00:06:56,726 This is because htop was developed back in 2004 187 00:06:56,726 --> 00:07:00,630 where top was developed all the way back in 1984. 188 00:07:00,630 --> 00:07:02,490 Now the one drawback is that htop 189 00:07:02,490 --> 00:07:05,700 is not installed by default on all Linux distributions. 190 00:07:05,700 --> 00:07:07,710 So if you want to use htop 191 00:07:07,710 --> 00:07:09,904 and it's not already installed on your distribution, 192 00:07:09,904 --> 00:07:12,270 you're going to have to install it using a package manager 193 00:07:12,270 --> 00:07:14,310 or compiling it from the source code 194 00:07:14,310 --> 00:07:15,886 and installing it on your Linux server. 195 00:07:15,886 --> 00:07:17,580 Now really, when it comes down 196 00:07:17,580 --> 00:07:20,250 to some of the big differences between top and htop, 197 00:07:20,250 --> 00:07:22,770 it really comes down to how it presents the data. 198 00:07:22,770 --> 00:07:23,670 When you're using top, 199 00:07:23,670 --> 00:07:26,040 it's a very basic black and white interface 200 00:07:26,040 --> 00:07:27,360 and it's text only. 201 00:07:27,360 --> 00:07:28,832 But when you start using htop, 202 00:07:28,832 --> 00:07:31,110 you're going to get a more colorful interface 203 00:07:31,110 --> 00:07:32,910 with some text-based graphics 204 00:07:32,910 --> 00:07:35,040 that'll depict exactly what's going on. 205 00:07:35,040 --> 00:07:38,227 For example, at the top of the screen in an htop display, 206 00:07:38,227 --> 00:07:41,273 you're going to see a graphical representation of the CPU, 207 00:07:41,273 --> 00:07:43,050 memory, and swap space 208 00:07:43,050 --> 00:07:45,720 and how much is being used and how much is free. 209 00:07:45,720 --> 00:07:47,958 Then, it will go on to show you all the different processes, 210 00:07:47,958 --> 00:07:49,782 the users that are running those processes, 211 00:07:49,782 --> 00:07:52,170 and all the information you're going to want to know 212 00:07:52,170 --> 00:07:54,357 about those processes on your given system, 213 00:07:54,357 --> 00:07:56,202 and it does this in a color-coded way, 214 00:07:56,202 --> 00:07:57,870 so it becomes really easy 215 00:07:57,870 --> 00:08:00,060 to quickly pick out the information you're looking for 216 00:08:00,060 --> 00:08:02,910 when you're using htop instead of top. 217 00:08:02,910 --> 00:08:06,543 Another command we want to talk about is systemd-analyze. 218 00:08:06,543 --> 00:08:08,970 Now the systemd-analyze command 219 00:08:08,970 --> 00:08:10,860 is used to retrieve performance statistics 220 00:08:10,860 --> 00:08:12,390 for boot operations. 221 00:08:12,390 --> 00:08:14,090 The command takes one or more sub commands 222 00:08:14,090 --> 00:08:17,100 that determine what type of information is going to be printed 223 00:08:17,100 --> 00:08:19,170 and how it's going to be printed. 224 00:08:19,170 --> 00:08:21,613 Now for process management and troubleshooting, 225 00:08:21,613 --> 00:08:24,439 blame is going to be the most relevant sub command. 226 00:08:24,439 --> 00:08:27,990 This sub command will print a list of all the systemd units 227 00:08:27,990 --> 00:08:29,580 that were executed at boot 228 00:08:29,580 --> 00:08:32,033 along with the time it took for each unit to execute. 229 00:08:32,033 --> 00:08:35,520 You can then use the systemd-analyze blame function 230 00:08:35,520 --> 00:08:37,679 to identify services and other units 231 00:08:37,679 --> 00:08:40,980 that make the system really slow during the bootup process 232 00:08:40,980 --> 00:08:43,076 To run the systemd-analyze command, 233 00:08:43,076 --> 00:08:46,290 simply type in systemd-analyze, 234 00:08:46,290 --> 00:08:48,540 the options and the sub command. 235 00:08:48,540 --> 00:08:51,570 The list of open files or lsof command 236 00:08:51,570 --> 00:08:53,337 is going to print out a list of all the files 237 00:08:53,337 --> 00:08:54,810 that are currently open 238 00:08:54,810 --> 00:08:57,276 to all your active processes on the system. 239 00:08:57,276 --> 00:08:59,420 This can include everything from a text file 240 00:08:59,420 --> 00:09:01,020 to a device file. 241 00:09:01,020 --> 00:09:03,240 Basically, any object on that system 242 00:09:03,240 --> 00:09:04,847 can be parsed as a file. 243 00:09:04,847 --> 00:09:06,674 Now sometimes, you're going to be prevented 244 00:09:06,674 --> 00:09:08,340 from modifying a file 245 00:09:08,340 --> 00:09:10,650 if it's already open by another process. 246 00:09:10,650 --> 00:09:12,882 So you can use the lsof command 247 00:09:12,882 --> 00:09:14,793 to identify the offending process 248 00:09:14,793 --> 00:09:17,711 that you need to terminate so you can take over that file. 249 00:09:17,711 --> 00:09:20,160 You can also use the lsof command 250 00:09:20,160 --> 00:09:22,317 to analyze how a process uses files 251 00:09:22,317 --> 00:09:23,730 which can be really helpful 252 00:09:23,730 --> 00:09:25,537 if you're identifying malicious processes 253 00:09:25,537 --> 00:09:28,033 or processes that have unwanted side effects 254 00:09:28,033 --> 00:09:29,591 in a given system. 255 00:09:29,591 --> 00:09:32,640 The lsof command or list of open files 256 00:09:32,640 --> 00:09:35,010 will print each file that's opened by a process 257 00:09:35,010 --> 00:09:37,020 to its own line on the screen. 258 00:09:37,020 --> 00:09:38,150 It's going to print information 259 00:09:38,150 --> 00:09:40,710 such as the name of the command or process, 260 00:09:40,710 --> 00:09:42,668 the PID, the invoking user, 261 00:09:42,668 --> 00:09:44,001 the file descriptor, 262 00:09:44,001 --> 00:09:46,980 and what kind of permissions the file has been opened with, 263 00:09:46,980 --> 00:09:49,740 as well as the type of file and the name of the file. 264 00:09:49,740 --> 00:09:50,910 To run this command, 265 00:09:50,910 --> 00:09:53,010 simply type in lsof 266 00:09:53,010 --> 00:09:55,489 and the options at the command prompt. 267 00:09:55,489 --> 00:09:58,470 Now another thing that the operating system does for us 268 00:09:58,470 --> 00:10:00,630 is it schedules out those processes 269 00:10:00,630 --> 00:10:03,180 to give each one some CPU time. 270 00:10:03,180 --> 00:10:05,070 This is known as the scheduler. 271 00:10:05,070 --> 00:10:07,320 Now the scheduler is usually really effective 272 00:10:07,320 --> 00:10:09,090 at assigning CPU time 273 00:10:09,090 --> 00:10:10,625 but not every decision it makes 274 00:10:10,625 --> 00:10:12,870 is going to be considered optimal. 275 00:10:12,870 --> 00:10:15,920 In some cases, you may need to manually tell your CPU 276 00:10:15,920 --> 00:10:18,893 to prioritize certain processes over others. 277 00:10:18,893 --> 00:10:20,877 Now processes are going to be prioritized 278 00:10:20,877 --> 00:10:24,030 based on a given number that's associated with it, 279 00:10:24,030 --> 00:10:26,790 anywhere from -20 to 19. 280 00:10:26,790 --> 00:10:30,466 This value is known as the nice value or the niceness value. 281 00:10:30,466 --> 00:10:32,220 If you have a lower number, 282 00:10:32,220 --> 00:10:34,560 it means you have a higher priority. 283 00:10:34,560 --> 00:10:36,630 For example, if you have a process 284 00:10:36,630 --> 00:10:38,940 with a nice value of -18, 285 00:10:38,940 --> 00:10:40,595 it's going to get more CPU time 286 00:10:40,595 --> 00:10:43,109 than a process with a nice value of 15 287 00:10:43,109 --> 00:10:47,010 because -18 is much smaller than 15. 288 00:10:47,010 --> 00:10:49,260 Now a process is going to inherit a nice value 289 00:10:49,260 --> 00:10:50,700 from its parent as well. 290 00:10:50,700 --> 00:10:54,300 And by default, the value it's always going to get is zero. 291 00:10:54,300 --> 00:10:56,580 The nice command will enable you to run a command 292 00:10:56,580 --> 00:10:58,860 with different nice values than the default one 293 00:10:58,860 --> 00:11:00,060 it's going to be getting. 294 00:11:00,060 --> 00:11:02,730 To do this, you're going to use the -n option, 295 00:11:02,730 --> 00:11:04,410 and then specify a number. 296 00:11:04,410 --> 00:11:06,420 If you don't specify an integer, 297 00:11:06,420 --> 00:11:09,570 it's going to go ahead and assume you wanted a 10 increment. 298 00:11:09,570 --> 00:11:11,354 Now by running nice without any options, 299 00:11:11,354 --> 00:11:14,049 you're going to see a default nice value. 300 00:11:14,049 --> 00:11:17,400 You must be the root user or have root user authority 301 00:11:17,400 --> 00:11:19,800 to run this command at a higher priority 302 00:11:19,800 --> 00:11:22,320 if you want to be able to change people's values. 303 00:11:22,320 --> 00:11:25,410 To use this, type in nice -n, 304 00:11:25,410 --> 00:11:27,240 the nice value increment you want 305 00:11:27,240 --> 00:11:29,430 and the command you want to change. 306 00:11:29,430 --> 00:11:31,470 Now the nice command is usually going to be used 307 00:11:31,470 --> 00:11:34,290 to start a new process at a given priority 308 00:11:34,290 --> 00:11:36,060 but there's another command out there 309 00:11:36,060 --> 00:11:38,490 if you want to change an existing process 310 00:11:38,490 --> 00:11:39,780 and its nice value, 311 00:11:39,780 --> 00:11:41,910 and this is known as renice, 312 00:11:41,910 --> 00:11:43,680 renice is the command to enable you 313 00:11:43,680 --> 00:11:45,210 to alter the scheduling priority 314 00:11:45,210 --> 00:11:47,460 of an already running process. 315 00:11:47,460 --> 00:11:49,320 You're going to use the renice command 316 00:11:49,320 --> 00:11:52,800 with the -n option to specify your new nice value 317 00:11:52,800 --> 00:11:54,870 that you want that process to have. 318 00:11:54,870 --> 00:11:57,060 Now if you want to, you can also renice 319 00:11:57,060 --> 00:12:00,012 not just a single process but an entire process group, 320 00:12:00,012 --> 00:12:02,910 and you do this using the -g option. 321 00:12:02,910 --> 00:12:05,783 This will cause all the processes in a given process group 322 00:12:05,783 --> 00:12:07,447 to have their nice values altered 323 00:12:07,447 --> 00:12:09,600 based on what you told it. 324 00:12:09,600 --> 00:12:11,143 Also, if you want to do a renice 325 00:12:11,143 --> 00:12:12,887 against everything with a user, 326 00:12:12,887 --> 00:12:16,200 you can do that too by using the -u option. 327 00:12:16,200 --> 00:12:18,900 This will alter the nice values of all the processes 328 00:12:18,900 --> 00:12:20,850 that are owned by a given user. 329 00:12:20,850 --> 00:12:22,080 To use renice, 330 00:12:22,080 --> 00:12:25,020 all you have to do is enter renice -n, 331 00:12:25,020 --> 00:12:26,601 the nice value, the options, 332 00:12:26,601 --> 00:12:29,587 and the identifier on the command line interface. 333 00:12:29,587 --> 00:12:31,890 Now when we're working in the command line, 334 00:12:31,890 --> 00:12:33,900 we're usually working in Bash, 335 00:12:33,900 --> 00:12:35,820 and with Bash, most of your commands 336 00:12:35,820 --> 00:12:38,610 are executed in the foreground by default. 337 00:12:38,610 --> 00:12:40,770 The problem with this is that your command prompt 338 00:12:40,770 --> 00:12:42,495 can become consumed and unusable 339 00:12:42,495 --> 00:12:45,087 until that command finishes processing. 340 00:12:45,087 --> 00:12:46,620 Now for simpler commands 341 00:12:46,620 --> 00:12:48,930 like the date command or the copy command, 342 00:12:48,930 --> 00:12:51,420 or the list command, this isn't really a problem, 343 00:12:51,420 --> 00:12:53,637 but if you're running a longer script or a longer command, 344 00:12:53,637 --> 00:12:56,244 this can start becoming a real issue for you 345 00:12:56,244 --> 00:12:58,713 and it can become possible where you have these commands 346 00:12:58,713 --> 00:13:00,870 that are just taking over your command prompt 347 00:13:00,870 --> 00:13:02,490 and you can't get a new command in 348 00:13:02,490 --> 00:13:04,530 because you're waiting for it too long. 349 00:13:04,530 --> 00:13:06,150 So one of the things you can do 350 00:13:06,150 --> 00:13:08,606 is you can actually run the commands in the background, 351 00:13:08,606 --> 00:13:11,040 and this way, you can execute them 352 00:13:11,040 --> 00:13:13,050 but not consume your shell. 353 00:13:13,050 --> 00:13:16,138 To do this, you can use the fg or foreground command 354 00:13:16,138 --> 00:13:18,298 or the bg, background command, 355 00:13:18,298 --> 00:13:19,690 and this will move the processes 356 00:13:19,690 --> 00:13:21,869 either into or out of view. 357 00:13:21,869 --> 00:13:24,180 Now there are a lot of different options you can use 358 00:13:24,180 --> 00:13:27,480 with both the foreground and background processes. 359 00:13:27,480 --> 00:13:30,419 The foreground or fg can be used with percent, 360 00:13:30,419 --> 00:13:34,380 and then a job ID to bring a job back to the foreground. 361 00:13:34,380 --> 00:13:35,790 If you use Ctrl + Z, 362 00:13:35,790 --> 00:13:37,336 this will halt a job temporarily 363 00:13:37,336 --> 00:13:39,300 so you can use the background command 364 00:13:39,300 --> 00:13:42,810 or bg and a percent sign with a job ID 365 00:13:42,810 --> 00:13:44,918 to push that job to the background. 366 00:13:44,918 --> 00:13:47,080 And if you use the ampersand sign, 367 00:13:47,080 --> 00:13:48,900 this will actually start the command 368 00:13:48,900 --> 00:13:50,130 running in the background 369 00:13:50,130 --> 00:13:52,260 when it's added at the end of the command. 370 00:13:52,260 --> 00:13:54,328 For example, if I use the date command, 371 00:13:54,328 --> 00:13:57,256 I could put date, ampersand and hit Enter, 372 00:13:57,256 --> 00:13:59,848 and it would run the date command in the background. 373 00:13:59,848 --> 00:14:02,280 Now the foreground and background command 374 00:14:02,280 --> 00:14:05,820 are actually part of a larger job control command function. 375 00:14:05,820 --> 00:14:08,036 The main function for this is known as jobs. 376 00:14:08,036 --> 00:14:10,020 If you enter jobs with a command prompt, 377 00:14:10,020 --> 00:14:12,120 it's going to list out all of those jobs you have 378 00:14:12,120 --> 00:14:14,400 either in the foreground or the background. 379 00:14:14,400 --> 00:14:15,690 This gives you a master list, 380 00:14:15,690 --> 00:14:17,313 and then you can then select which ones you want 381 00:14:17,313 --> 00:14:20,782 by using the fg or bg commands to bring them forward. 382 00:14:20,782 --> 00:14:22,950 If you want to stop a foreground job 383 00:14:22,950 --> 00:14:25,020 and place it in the background when you're working in it, 384 00:14:25,020 --> 00:14:27,690 you can simply type in Ctrl + Z. 385 00:14:27,690 --> 00:14:29,880 If you're working in the command line environment 386 00:14:29,880 --> 00:14:31,731 and you're running a program that you want to stop, 387 00:14:31,731 --> 00:14:34,890 you can force quit it by using Ctrl + C 388 00:14:34,890 --> 00:14:36,630 at the command line. 389 00:14:36,630 --> 00:14:38,970 If you want to log out of your current user session, 390 00:14:38,970 --> 00:14:41,730 you can do that by using Ctrl + D. 391 00:14:41,730 --> 00:14:42,930 All of these are shortcuts 392 00:14:42,930 --> 00:14:44,100 for you to be able to go ahead 393 00:14:44,100 --> 00:14:45,872 and move in or out of a process 394 00:14:45,872 --> 00:14:48,420 as you need to when you're working in the jobs 395 00:14:48,420 --> 00:14:50,400 or the command line interface. 396 00:14:50,400 --> 00:14:52,620 Now some other commands we have that are important 397 00:14:52,620 --> 00:14:56,190 are things like nohup and no hangup. 398 00:14:56,190 --> 00:14:58,222 Now nohup or no hangup 399 00:14:58,222 --> 00:15:00,073 is a command that's going to prevent a process 400 00:15:00,073 --> 00:15:02,649 from ending when the user logs off. 401 00:15:02,649 --> 00:15:04,890 For example, let's say an administrator 402 00:15:04,890 --> 00:15:06,180 launched a backup script, 403 00:15:06,180 --> 00:15:07,542 and then logs off the system. 404 00:15:07,542 --> 00:15:10,830 By default, that script would actually stop running. 405 00:15:10,830 --> 00:15:12,510 That would be bad because it means our backup 406 00:15:12,510 --> 00:15:14,040 is going to stop running. 407 00:15:14,040 --> 00:15:17,940 But if we use nohup in front of the normal command, 408 00:15:17,940 --> 00:15:21,210 like nohup.backup script, 409 00:15:21,210 --> 00:15:22,710 that script will continue to run 410 00:15:22,710 --> 00:15:25,020 even after the administrator logged off. 411 00:15:25,020 --> 00:15:26,250 So to run this, 412 00:15:26,250 --> 00:15:30,600 simply type in nohup and the command or script name. 413 00:15:30,600 --> 00:15:32,130 Different commands are also going to be used 414 00:15:32,130 --> 00:15:33,560 to send signals to processes 415 00:15:33,560 --> 00:15:35,250 to either terminate or kill them 416 00:15:35,250 --> 00:15:37,050 when you don't need them anymore. 417 00:15:37,050 --> 00:15:39,150 This is necessary when a process hangs, 418 00:15:39,150 --> 00:15:41,460 which causes a system to become unstable, 419 00:15:41,460 --> 00:15:44,400 or it fails to relinquish control over a file 420 00:15:44,400 --> 00:15:46,140 that you're trying to modify. 421 00:15:46,140 --> 00:15:48,900 A simple way to get this taken care of is by issuing 422 00:15:48,900 --> 00:15:50,160 the kill command. 423 00:15:50,160 --> 00:15:52,748 The kill command is going to send any specified signal, 424 00:15:52,748 --> 00:15:55,470 or by default determination signal, 425 00:15:55,470 --> 00:15:57,330 to one or more processes. 426 00:15:57,330 --> 00:15:58,892 To run this, type in kill, 427 00:15:58,892 --> 00:16:01,379 the options and the process ID. 428 00:16:01,379 --> 00:16:04,003 Another variant of this is known as pkill. 429 00:16:04,003 --> 00:16:07,340 Now the pkill command is going to send a specified signal, 430 00:16:07,340 --> 00:16:09,500 or by default determination signal, 431 00:16:09,500 --> 00:16:11,734 to a process based on a pattern. 432 00:16:11,734 --> 00:16:15,150 Now for example, let's say you start top in one terminal, 433 00:16:15,150 --> 00:16:17,590 and then you issue pkill top in another terminal, 434 00:16:17,590 --> 00:16:20,250 you're going to see that top terminates. 435 00:16:20,250 --> 00:16:24,120 This is because the word top match the pattern of T-O-P 436 00:16:24,120 --> 00:16:25,740 or the top process. 437 00:16:25,740 --> 00:16:27,057 So if you want to run a command like this, 438 00:16:27,057 --> 00:16:29,460 simply enter pkill, the options, 439 00:16:29,460 --> 00:16:31,410 and the pattern you want to terminate. 440 00:16:31,410 --> 00:16:33,236 On the other hand, the killall command 441 00:16:33,236 --> 00:16:35,134 will send any specified signal, 442 00:16:35,134 --> 00:16:37,676 or again by default determination signal, 443 00:16:37,676 --> 00:16:41,520 to all processes that are matching the name specified. 444 00:16:41,520 --> 00:16:42,510 To run this command, 445 00:16:42,510 --> 00:16:44,130 simply type in killall, 446 00:16:44,130 --> 00:16:46,620 the options and the process name. 447 00:16:46,620 --> 00:16:48,229 Now you can also use the kill command 448 00:16:48,229 --> 00:16:50,877 with the process table to end processes 449 00:16:50,877 --> 00:16:52,860 and you can do this by entering kill, 450 00:16:52,860 --> 00:16:54,480 followed by the PID 451 00:16:54,480 --> 00:16:56,940 for that specific process that you want to terminate. 452 00:16:56,940 --> 00:17:00,073 However, the process table may display processes 453 00:17:00,073 --> 00:17:01,860 that do not belong to you. 454 00:17:01,860 --> 00:17:05,278 And as a user, you can only kill things that you own. 455 00:17:05,278 --> 00:17:07,410 As root though, you can kill 456 00:17:07,410 --> 00:17:09,358 any process you want on the system 457 00:17:09,358 --> 00:17:11,579 and there are many different ways to kill a process, 458 00:17:11,579 --> 00:17:13,230 like I just mentioned, 459 00:17:13,230 --> 00:17:15,720 and each one of them will be mapped to a signal. 460 00:17:15,720 --> 00:17:19,140 Now the signal determines how to actually kill your process. 461 00:17:19,140 --> 00:17:21,354 By default, we use the termination signal. 462 00:17:21,354 --> 00:17:24,458 For example, if you want to terminate a process gracefully, 463 00:17:24,458 --> 00:17:27,491 you can give it time to clean up any last minute activities. 464 00:17:27,491 --> 00:17:30,611 However, some processes will ignore this signal 465 00:17:30,611 --> 00:17:32,490 or they'll simply fail to terminate 466 00:17:32,490 --> 00:17:34,320 in a reasonable amount of time. 467 00:17:34,320 --> 00:17:35,670 For those processes, 468 00:17:35,670 --> 00:17:38,040 you may have to issue a more aggressive signal. 469 00:17:38,040 --> 00:17:39,270 Now there are many different types 470 00:17:39,270 --> 00:17:40,980 of kill signals out there. 471 00:17:40,980 --> 00:17:42,390 Each signal has a name 472 00:17:42,390 --> 00:17:44,880 and one or more corresponding number values 473 00:17:44,880 --> 00:17:46,950 and you can use them with the kill command. 474 00:17:46,950 --> 00:17:49,770 The SIGHUP signal has a value of one, 475 00:17:49,770 --> 00:17:51,030 and it's sent to a process 476 00:17:51,030 --> 00:17:53,280 when its controlling terminal is closed. 477 00:17:53,280 --> 00:17:55,290 If that terminal is closed or disconnected, 478 00:17:55,290 --> 00:17:56,970 it's said to have hung up, 479 00:17:56,970 --> 00:17:59,250 which is why this signal is called SIGHUP, 480 00:17:59,250 --> 00:18:03,690 or S-I-G-H-U-P, which stands for signal hang up. 481 00:18:03,690 --> 00:18:06,510 This SIGINT signal is going to have a value of two 482 00:18:06,510 --> 00:18:08,670 that interrupts a process from the terminal 483 00:18:08,670 --> 00:18:10,920 and it enables it to end gracefully. 484 00:18:10,920 --> 00:18:13,140 The signal can be caught or ignored. 485 00:18:13,140 --> 00:18:14,520 This is basically the same thing 486 00:18:14,520 --> 00:18:16,710 as pressing Ctrl + C at a terminal. 487 00:18:16,710 --> 00:18:18,957 A process might change a shortcut behavior over, 488 00:18:18,957 --> 00:18:21,840 and in those cases, you can use SIGINT 489 00:18:21,840 --> 00:18:25,440 Now, the SIGKILL is going to be a signal with a value of nine 490 00:18:25,440 --> 00:18:27,570 and it kills the process immediately. 491 00:18:27,570 --> 00:18:30,090 The signal cannot be caught or ignored. 492 00:18:30,090 --> 00:18:32,940 This is typically used as a last resort. 493 00:18:32,940 --> 00:18:35,220 Another one is SIGTERM. 494 00:18:35,220 --> 00:18:37,560 This is a signal with a value of 15 495 00:18:37,560 --> 00:18:39,030 and it terminates a process 496 00:18:39,030 --> 00:18:41,040 and enables it to end gracefully. 497 00:18:41,040 --> 00:18:43,440 This signal can be caught or ignored 498 00:18:43,440 --> 00:18:45,450 and is typically sent by the kill command 499 00:18:45,450 --> 00:18:47,880 or another program and not from the terminal. 500 00:18:47,880 --> 00:18:50,970 The SIGSTOP signal is going to have a value 501 00:18:50,970 --> 00:18:53,760 of 17, 19 and 23, 502 00:18:53,760 --> 00:18:55,710 and it pauses a process. 503 00:18:55,710 --> 00:18:58,890 This signal cannot be caught or ignored by the program 504 00:18:58,890 --> 00:19:01,890 and this is typically sent from kill or another program 505 00:19:01,890 --> 00:19:04,080 and not by the terminal itself. 506 00:19:04,080 --> 00:19:08,940 The SIGSTP signal has a value of 18, 20 and 24, 507 00:19:08,940 --> 00:19:11,550 and it pauses a process from the terminal. 508 00:19:11,550 --> 00:19:14,460 This signal can be caught and it can be ignored. 509 00:19:14,460 --> 00:19:17,490 This is the same as pressing Ctrl + Z at your terminal 510 00:19:17,490 --> 00:19:20,400 and a process might change this shortcut behavior, however, 511 00:19:20,400 --> 00:19:21,677 in which case you would use 512 00:19:21,677 --> 00:19:25,053 the sigstop instead or SIGSTP. 513 00:19:25,920 --> 00:19:27,810 So let me give you a couple of examples 514 00:19:27,810 --> 00:19:29,400 to show you how this works. 515 00:19:29,400 --> 00:19:33,900 Let's say I had a process that had a process ID of 921. 516 00:19:33,900 --> 00:19:35,370 Now if I wanted to kill that, 517 00:19:35,370 --> 00:19:36,657 I could send the kill signal 518 00:19:36,657 --> 00:19:41,520 by using kill 15 921 at the command prompt. 519 00:19:41,520 --> 00:19:44,550 And this would then send the signal of 15 520 00:19:44,550 --> 00:19:47,130 to kill process 921 521 00:19:47,130 --> 00:19:49,252 which is the signal termination signal. 522 00:19:49,252 --> 00:19:51,570 Now if I wanted to use signal 17, 523 00:19:51,570 --> 00:19:54,690 I could enter kill 17 921 524 00:19:54,690 --> 00:19:56,580 and this would pause that process 525 00:19:56,580 --> 00:19:59,040 instead of removing the process entirely. 526 00:19:59,040 --> 00:20:00,210 Now most of the time, 527 00:20:00,210 --> 00:20:03,090 people who use Linux prefer to do their own troubleshooting 528 00:20:03,090 --> 00:20:06,060 and solving problems becomes an art and a science. 529 00:20:06,060 --> 00:20:07,500 Solving technical problems, 530 00:20:07,500 --> 00:20:09,300 such as those that occur with computers 531 00:20:09,300 --> 00:20:12,300 requires a great deal of specialized knowledge too. 532 00:20:12,300 --> 00:20:13,350 We may use guidelines 533 00:20:13,350 --> 00:20:15,180 when troubleshooting our process issues 534 00:20:15,180 --> 00:20:16,350 and troubleshooting begins 535 00:20:16,350 --> 00:20:17,937 with the identification of a problem 536 00:20:17,937 --> 00:20:20,850 and it doesn't end until the services have been restored 537 00:20:20,850 --> 00:20:24,450 and the problem no longer is adversely affecting our users. 538 00:20:24,450 --> 00:20:27,270 Remember, you want to gather information about a process 539 00:20:27,270 --> 00:20:30,900 including its process ID or PID and the state 540 00:20:30,900 --> 00:20:33,480 when you start to troubleshoot these process issues. 541 00:20:33,480 --> 00:20:35,970 You can use the ps command to print information 542 00:20:35,970 --> 00:20:37,680 about all your running processes 543 00:20:37,680 --> 00:20:39,780 including CPU utilization. 544 00:20:39,780 --> 00:20:41,275 For example, if you have a process 545 00:20:41,275 --> 00:20:43,980 that's consuming over 50% of the CPU, 546 00:20:43,980 --> 00:20:47,220 it may cause performance issues across the entire system 547 00:20:47,220 --> 00:20:49,290 and it becomes worth investigating. 548 00:20:49,290 --> 00:20:51,000 If you're using the top command, 549 00:20:51,000 --> 00:20:53,130 you can retrieve a dynamic and interactive list 550 00:20:53,130 --> 00:20:54,990 of all your running processes. 551 00:20:54,990 --> 00:20:57,870 If you use the systemd-analyze blame command, 552 00:20:57,870 --> 00:20:59,940 you can determine what startup processes 553 00:20:59,940 --> 00:21:02,160 are slowing down the boot operations. 554 00:21:02,160 --> 00:21:03,840 If you use lsof, 555 00:21:03,840 --> 00:21:05,580 you can identify which processes 556 00:21:05,580 --> 00:21:08,070 have files that are open right now and in use. 557 00:21:08,070 --> 00:21:09,960 If you use nice and renice, 558 00:21:09,960 --> 00:21:12,990 you can prioritize certain processes over others. 559 00:21:12,990 --> 00:21:14,970 If you use fg and bg, 560 00:21:14,970 --> 00:21:18,030 you can measure foreground and background processes. 561 00:21:18,030 --> 00:21:19,890 And if you use nohup, 562 00:21:19,890 --> 00:21:23,190 you can keep a command running even after you log off. 563 00:21:23,190 --> 00:21:26,490 Finally, if you use kill and its associated commands, 564 00:21:26,490 --> 00:21:28,440 you can terminate problem processes 565 00:21:28,440 --> 00:21:30,690 and restore the system disability. 566 00:21:30,690 --> 00:21:32,340 Each time you solve a problem, 567 00:21:32,340 --> 00:21:34,320 you're going to increase your troubleshooting skills 568 00:21:34,320 --> 00:21:35,910 by gaining more experience. 569 00:21:35,910 --> 00:21:36,960 And as you start to learn 570 00:21:36,960 --> 00:21:39,690 how to combine these different steps or skip steps, 571 00:21:39,690 --> 00:21:42,290 you can start reaching a solution much more quickly.