Processes in Linux¶
A process is a program (think about any Linux application) in a running state. It contains various details, like the memory space the program needs, a process id, the files opened by the process, etc.
How to view all running processes?¶
The following command shows all the processes from your computer.
$ ps aux
USER PID %CPU %MEM VSZ RSS TTY STAT START TIME COMMAND
root 1 0.0 0.0 215356 4984 ? Ss May29 0:28 /usr/lib/systemd/systemd --system --deserialize 19
root 2 0.0 0.0 0 0 ? S May29 0:00 [kthreadd]
root 4 0.0 0.0 0 0 ? S< May29 0:00 [kworker/0:0H]
root 6 0.0 0.0 0 0 ? S May29 0:11 [ksoftirqd/0]
root 7 0.0 0.0 0 0 ? S May29 8:27 [rcu_sched]
... long output
You can see that the output also tells you under which user the process is running, what the actual command being used is, and the percentage of CPU and memory usage.
The PID column shows the process id; you can see that the systemd process has PID 1, which means it is the first process to start in the system.
How to find a particular process?¶
Let’s say, I want to know the process id of the Firefox browser in my system. I can use the following command to find that information.
$ ps aux | grep firefox
kdas 26752 96.1 9.7 2770724 763436 ? Sl 16:16 0:35 /usr/lib64/firefox/firefox
kdas 26919 0.0 0.0 118520 980 pts/3 S+ 16:17 0:00 grep --color=auto firefox
Here, we are first running the ps command, and then passing the output of that to the next command using the | character. In this case, as you see, grep is that second command. We can find and look for text using the grep tool. We will learn more about grep in the future.
How to kill/stop a particular process?¶
We can kill/stop any process using the kill command. We found out, in the last example, that the id of the Firefox process in my computer is 26752, we can use that id to kill it.
$ kill 26752
If there is no error message, you’ll find that Firefox has disappeared.
Finding out list of open files¶
lsof command will show list of all open files. The man page has more details about the different command line options available.
Signals¶
Signals are a limited way to communicate to a process. You can think about them as notifications to a process, and depending on the signal handler in the code, the process does something with that signal. The kill command actually sends a signal to the given process id, the default signal is TERM, which says to terminate the process. To directly/forcibly kill a process, you can send the KILL signal.
$ kill -9 26752
Here 9 is number representation of the KILL signal. To know more about Linux signals, read the man page.
$ man 7 signal
kill command also has a -l flag, which prints all of the signal names, and numbers on the screen.
$ 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
top command¶
top is a very useful command while using a Linux system. It’s a quick way to know about all the running processes in the system, and their related status about CPU and memory usage in general. To get out of top, press the key q.
top - 17:37:28 up 24 days, 11:52, 2 users, load average: 0.57, 0.73, 0.75
Tasks: 372 total, 2 running, 370 sleeping, 0 stopped, 0 zombie
%Cpu(s): 11.6 us, 2.6 sy, 0.0 ni, 84.9 id, 0.1 wa, 0.3 hi, 0.5 si, 0.0 st
KiB Mem : 7858752 total, 1701052 free, 4444136 used, 1713564 buff/cache
KiB Swap: 3268604 total, 1558396 free, 1710208 used. 2431656 avail Mem
PID USER PR NI VIRT RES SHR S %CPU %MEM TIME+ COMMAND
28300 kdas 20 0 1502016 287340 44396 R 25.0 3.7 290:56.60 chrome
2668 kdas 9 -11 2067292 9756 7164 S 6.2 0.1 166:06.48 pulseaudio
15122 kdas 20 0 771844 33104 11352 S 6.2 0.4 39:24.60 gnome-terminal-
24760 kdas 20 0 1945840 209128 76952 S 6.2 2.7 1:41.15 code
27526 kdas 20 0 156076 4268 3516 R 6.2 0.1 0:00.01 top
1 root 20 0 215356 4880 3108 S 0.0 0.1 0:28.25 systemd
2 root 20 0 0 0 0 S 0.0 0.0 0:00.66 kthreadd
4 root 0 -20 0 0 0 S 0.0 0.0 0:00.00 kworker/0:0H
6 root 20 0 0 0 0 S 0.0 0.0 0:11.79 ksoftirqd/0
7 root 20 0 0 0 0 S 0.0 0.0 8:28.06 rcu_sched
... long output
By the way, feel free to press 1 and see if anything changes in the top command output.
Load average¶
If you look at the top output carefully, you will find load average mentioned. Actually, there are 3 numbers provided; these are the load averages of the system in the last one minute, 5 minutes ago, and 15 minutes ago.
load average: 0.57, 0.73, 0.75
In simple words, load average means the average time any process has to wait to get access to the CPU (or other resources), in idle state the load average is 0. This information is a quick way to learn about the system, if the system is slow to respond, just looking at the load-average, and then the rest of the top output should be a good starting point.
htop tool¶
htop is a modern version of the top tool. It has many more features, interactiveness being the biggest amongst them. htop does not come by default in most of the Linux installations, which means you will have to install it using the system’s package management tool.
These are the ways to install it in Fedora and in Debian/Ubuntu
$ sudo dnf install htop -y
$ sudo apt-get install htop
To know more about htop, please read the man page.
$ man htop
More about Linux processes¶
You can learn more about Linux processes in the glibc manual. Use the info command to find out more.
$ info libc process
/proc directory¶
/proc is a special directory in our filesystem. This is a virtual filesystem which contains information about all the running processes, and information about the hardware present in the system. You will find that the files in the virtual filesystem are 0 in size.
Now we’ll learn about a few files inside this directory.
/proc/cpuinfo¶
/proc/cpuinfo file has information about the CPU in your system. It includes the model number, and also the various flags available in that particular CPU model.
/proc/cmdline¶
/proc/cmdline file has all the parameters passed to the kernel at the bootup time. The following is a cloud-based virtual machine.
$ cat /proc/cmdline
BOOT_IMAGE=/boot/vmlinuz-4.8.6-300.fc25.x86_64 root=UUID=9be70055-35f2-4a57-b120-5a003dfdb504 ro no_timer_check console=tty1 console=ttyS0,115200n8 rhgb quiet console=ttyS1 LANG=en_US.UTF-8 initrd=/boot/initramfs-4.8.6-300.fc25.x86_64.img
/proc/meminfo¶
/proc/meminfo contains information related to the memory in the system. You can see the total amount RAM, the available memory and other values there.
$ cat /proc/meminfo
MemTotal: 4046820 kB
MemFree: 2960568 kB
MemAvailable: 3696216 kB
Buffers: 53756 kB
Cached: 830052 kB
SwapCached: 0 kB
Active: 347216 kB
Inactive: 575692 kB
Active(anon): 39388 kB
Inactive(anon): 196 kB
Active(file): 307828 kB
Inactive(file): 575496 kB
Unevictable: 0 kB
Mlocked: 0 kB
SwapTotal: 0 kB
SwapFree: 0 kB
Dirty: 4 kB
Writeback: 0 kB
AnonPages: 39120 kB
Mapped: 42032 kB
Shmem: 488 kB
Slab: 141692 kB
SReclaimable: 114996 kB
SUnreclaim: 26696 kB
KernelStack: 1360 kB
PageTables: 2700 kB
NFS_Unstable: 0 kB
Bounce: 0 kB
WritebackTmp: 0 kB
CommitLimit: 2023408 kB
Committed_AS: 127752 kB
VmallocTotal: 34359738367 kB
VmallocUsed: 0 kB
VmallocChunk: 0 kB
HardwareCorrupted: 0 kB
AnonHugePages: 0 kB
ShmemHugePages: 0 kB
ShmemPmdMapped: 0 kB
CmaTotal: 0 kB
CmaFree: 0 kB
HugePages_Total: 0
HugePages_Free: 0
HugePages_Rsvd: 0
HugePages_Surp: 0
Hugepagesize: 2048 kB
DirectMap4k: 61296 kB
DirectMap2M: 4132864 kB
DirectMap1G: 2097152 kB
/proc/uptime¶
$ cat /proc/uptime
52820.32 104802.84
The first value in this file shows the number of seconds the system is up. The second value is the total number of idle seconds for each CPU, so for the modern systems, this value can be more than the first value.
/proc/sys/ & sysctl command¶
This directory is a special one for system administrators. This not only provides information, but also allows you to quickly change (enable/disable) different kernel features.
We use the sysctl command to view or edit the values for /proc/sys/. If you want to see all the different settings, use the following command.
$ sudo sysctl -a
[sudo] password for kdas:
abi.vsyscall32 = 1
crypto.fips_enabled = 0
debug.exception-trace = 1
debug.kprobes-optimization = 1
dev.cdrom.autoclose = 1
dev.cdrom.autoeject = 0
dev.cdrom.check_media = 0
dev.cdrom.debug = 0
dev.cdrom.info = CD-ROM information, Id: cdrom.c 3.20 2003/12/17
... long output
Enabling IP forward with sysctl¶
To enable IP forwarding to the VM(s), use the following command.
$ sudo sysctl -w net.ipv4.ip_forward=1
To check the current value, use the following command.
$ sysctl net.ipv4.ip_forward
You can see the same value in the /proc/sys/net/ipv4/ip_forward file too.
$ cat /proc/sys/net/ipv4/ip_forward
1
To make the change permanent, write the following in the /etc/sysctl.conf file.
net.ipv4.ip_forward = 1
Then, enable the changes using the following command.
$ sudo sysctl -p /etc/sysctl.conf
