What process is writing to file linux

Find Which Linux Process is Using Your Files or Ports

Find Which Linux Process is Using Your Files or Ports

If you have ever tried to unmount a filesystem, USB stick, cdrom, etc… but found that it was busy and therefore could not be unmounted, you will like the fuser command. Fuser’s man page tells us that fuser will “Show which processes use the named files, sockets, or filesystems.” But you can do more than that! You can also use fuser to send signals to your processes telling them to shutdown, terminate, pause, etc…

Use Fuser to Find Which Processes are Using a Filesystem

One of the things that used to annoy me the most was when I would try to unmount a filesystem and I would find that a process was using it and I would get the “device is busy” message.

Now, when I get that message, I simply use fuser to find which processes are using files in that filesystem.

Example: Show linux processes using /myfilesystem

As we look at the output from the fuser command above, we see that both the root user and the mary user have processes in the filesystem. We can review the columns and get lots of information. The USER, PID, and COMMAND columns are similar to the columns from the ps command and are self-explanatory. The USER column tells us the user id for the process. The PID column tells us the process id of that process. The COMMAND column tells us the command the process is running as well as the userid again in parenthesis. The ACCESS column tells us how the process is using the file. The ACCESS column may include he following letters referencing these access types:

ACCESS Column Code	Code’s Definition
F	file is open for writing.
f,o	The process has an open file.
r	The process’ root or home directory is on this filesystem (chroot).
c	The process’ current directory is on this filesystem.
e,t	The process is executing a file.
m,s	The process has a mapped file or is using a shared library.

Using LSOF (List Open Files)

We can also use the “lsof” (list open files) command to show a similar report.

Example: Using lsof to find processes using a specified filesystem.

Find and Killing Processes Using Fuser

For the more confident system administrators, fuser has a kill option (-k) that can be used to kill whatever processes are found using a file or filesystem.

Example: Find whatever processes are using the /myfilesystem and shut them down!

Interestingly, in the above example, we found that our very process was using the /myfilesystem filesystem. We ran the fuser command with the kill option (-k) telling it to kill the processes. As ours was amongst them, it killed our process.

Use Fuser to Find Which Processes / Daemons are Listening on a Port

Fuser has additional uses. Sometimes, we want to find out what process and userid are utilizing a port on our system. We can specify the IP version with -v4 for IPv4 or -v6 for IPv6.

Example: Find out what process is handling our http traffic on port 80

We were able to find out that varnishd is handling our http traffic on port 80/tcp.

Kill Whatever Process is Listening on a Port

Let’s imagine that you need to shutdown whatever is running on port 8001/tcp for some reason. Perhaps, you have a process that normally runs on that port, but something else has taken it while you have been performing some maintenance. We can use fuser to find and kill the process in one command by using the fuser command with the -k option.

Example: Find and kill whatever process is listening on port 8001/tcp

Specifying the Signal Fuser Will Use as it Kills a Process

You can change the signal used when killing a process using fuser by simply listing it as an option to your fuser command. You can list the available signals using “fuser -l”.

and then use withever signal you want it to use:

Other Resources

Fuser is a powerful tool that can help you find which linux process is using your files or ports. Here are a few other resources that can give you additional information on fuser.

Источник

How to check running process in Linux using command line

I am a new system administrator for the Linux operating system. How do I check running process in Linux using the command line option?

Tutorial details
Difficulty level	Easy
Root privileges	Yes
Requirements	Linux terminal
Est. reading time	4 mintues

One can use the Linux command line or terminal app to display a running process, change their priorities level, delete process and more. This page shows how to use various commands to list, kill and manage process on Linux.

Check running process in Linux

The procedure to monitor the running process in Linux using the command line is as follows:

1. Open the terminal window on Linux
2. For remote Linux server use the ssh command for log in purpose
3. Type the ps aux command to see all running process in Linux
4. Alternatively, you can issue the top command or htop command to view running process in Linux

Let us see some example and usage in details.

Please note that vivek@nixcraft:

$ is my shell prompt. You need to type commands after the $ prompt.

How to manage processes from the Linux terminal

The ps command is a traditional Linux command to lists running processes. The following command shows all processes running on your Linux based server or system:
vivek@nixcraft:

$ ps -aux
vivek@nixcraft:

1. root – User name
2. 1 – PID (Linux process ID)
3. 19:10 – Process start time
4. /sbin/init splash – Actual process or command

There may be too many processes. Hence, it uses the following less command/more command as pipe to display process one screen at a time:
vivek@nixcraft:

$ ps -aux | more
vivek@nixcraft:

$ sudo ps -aux | less
Press q to exit from above Linux pagers. You can search for a particular Linux process using grep command/egrep command:
vivek@nixcraft:

$ ps aux | grep firefox
vivek@nixcraft:

$ sudo ps aux | grep vim
vivek@nixcraft:

$ sudo ps -aux | egrep ‘sshd|openvpn|nginx’

• No ads and tracking
• In-depth guides for developers and sysadmins at Opensourceflare✨
• Join my Patreon to support independent content creators and start reading latest guides:
  • How to set up Redis sentinel cluster on Ubuntu or Debian Linux
  • How To Set Up SSH Keys With YubiKey as two-factor authentication (U2F/FIDO2)
  • How to set up Mariadb Galera cluster on Ubuntu or Debian Linux
  • A podman tutorial for beginners – part I (run Linux containers without Docker and in daemonless mode)
  • How to protect Linux against rogue USB devices using USBGuard

Join Patreon ➔

Linux pgrep command

Many variants of Linux comes with the pgrep command to search/find process. The syntax is:
vivek@nixcraft:

$ sudo pgrep sshd
vivek@nixcraft:

$ pgrep vim
vivek@nixcraft:

$ pgrep firefox
vivek@nixcraft:

Linux top command

The top command is another highly recommended method to see your Linux servers resource usage. One can see a list of top process that using the most memory or CPU or disk.
vivek@nixcraft:

$ sudo top
vivek@nixcraft:

Linux htop command to check running process in Linux

The htop command is an interactive process viewer and recommended method for Linux users. One can see a list of top process that using the most memory or CPU or disk and more:
vivek@nixcraft:

$ sudo htop
vivek@nixcraft:

Linux kill command

Want to kill a process? Try kill command. The syntax is:
vivek@nixcraft:

$ kill pid
vivek@nixcraft:

$ kill -signal pid
Find PID using ps, pgrep or top commands. Say you want to kill a PID # 16750, run:
vivek@nixcraft:

$ kill 16750
For some reason if the process can not be killed, try forceful killing:
vivek@nixcraft:

$ kill -9 16750
OR
vivek@nixcraft:

$ kill -KILL 16750

Linux pkill command

If you wish to kill a process by name, try pkill command. The syntax is:
vivek@nixcraft:

$ pkill processName
vivek@nixcraft:

$ pkill vim
vivek@nixcraft:

$ pkill firefox
vivek@nixcraft:

$ pkill -9 emacs
vivek@nixcraft:

$ sudo pkill -KILL php7-fpm

Linux killall command

The killall command kills processes by name, as opposed to the selection by PID as done by kill command:
vivek@nixcraft:

$ killall vim
vivek@nixcraft:

$ killall -9 emacs

Linux nice and renice command

The primary purpose of the nice command is to run a process/command at a lower or higher priority. Use the renice command to alter the nice value of one or more running Linux processes. The nice value can range from -20 to 19, with 19 being the lowest priority. Say, you want to compile software on a busy Linux server. You can set a very low priority, enter:
vivek@nixcraft:

$ nice -n 13 cc -c *.c &
Set a very high priority for a kernel update. Before rebooting Linux server, run:

Источник

Linux / UNIX List Open Files for Process

H ow do I list all open files for a Linux or UNIX process using command line options? How can I show open files per process under Linux?

Both Linux and Unix-like operating systems come with various utilities to find out open files associated with the process.

Tutorial details
Difficulty level	Easy
Root privileges	Yes
Requirements	None
Est. reading time	3 minutes

UNIX List Open Files For Process

First use the ps command command to get PID of process, enter:
$ ps -aef | grep $ ps -aef | grep httpd
Next pass this PID to pfiles command under Solaris Unix:
$ pfiles
$ pfiles 3533
See pfiles command documentation> for more information or type the following man command:
% man pfiles

FreeBSD list open files per process

On FreeBSD use the fstat command along with the ps command:
# ps aux | grep -i openvpn # filter outputs using the grep command #
# fstat -p
# fstat -p 1219
We can count open files count for openvpn process as follows using the wc command:
# fstat -p 1219 | grep -v ^USER | wc -l
The -p option passed to the fstat to report all files open by the specified process.

FreeBSD pstat command in action

Linux List Open Files For Process

First you need to find out PID of process. Simply use any one of the following command to obtain process id:
# ps aux | grep OR
$ ps -C -o pid=
For example, find out PID of firefox web-browser, enter:
$ ps -C firefox -o pid=
Output:

To list opne files for firefox process, enter:
$ ls -l /proc/7857/fd
Sample output:

• No ads and tracking
• In-depth guides for developers and sysadmins at Opensourceflare✨
• Join my Patreon to support independent content creators and start reading latest guides:
  • How to set up Redis sentinel cluster on Ubuntu or Debian Linux
  • How To Set Up SSH Keys With YubiKey as two-factor authentication (U2F/FIDO2)
  • How to set up Mariadb Galera cluster on Ubuntu or Debian Linux
  • A podman tutorial for beginners – part I (run Linux containers without Docker and in daemonless mode)
  • How to protect Linux against rogue USB devices using USBGuard

Join Patreon ➔

For privileged process use the sudo command and to count open files use the wc command on Linux as follows:
# Get process pid
sudo ps -C Xorg -o pid
sudo ls -l /proc/$/fd
# Say pid is 9497 for Xorg, then
sudo ls -l /proc/9497/fd | wc -l
We can use bash for loop as follows too:

Listing Open Files on Linux

Using lsof to display the processes using the most file handles

The lsof command list open files under all Linux distributions or UNIX-like operating system. Type the following command to list open file for process ID 351:
$ lsof -p 351
In this example display and count all open files for top 10 processes on Linux operating systems or server:
# lsof | awk ‘‘ | sort | uniq -c | sort -r | head
## force numeric sort by passing the ‘-n’ option to the sort ##
# lsof | awk ‘‘ | sort | uniq -c | sort -r -n | head

• lsof – Run the lsof to display all open files and send output to the awk
• awk ‘‘ – Display first field i.e. process name only
• uniq -c – Omit duplicate lines while prefix lines by the number of occurrences
• sort -r – Reverse sort
• head – Display top 10 process along with open files count

Conclusion

Now you know how to find open files per process on Linux, FreeBSD, and Unix-like systems using various command-line options. See how to increase the system-wide/user-wide number of available (open) file handles on Linux for more information.

🐧 Get the latest tutorials on Linux, Open Source & DevOps via

Источник

How to write the output into the file in Linux

How do I save terminal output to a file?

A command can receive input from a file and send output to a file.

Writing the output into the file

The syntax is
command > filename
For example, send output of the ls command to file named foo.txt
$ ls > foo.txt
View foo.txt using the cat command:
$ cat foo.txt
Please note that when you type ‘ls > foo.txt’, shell redirects the output of the ls command to a file named foo.txt, replacing the existing contents of the file. In other words, the contents of the file will be overwritten.

Appending the output or data to the file

The syntax is
command >> filename
For example the following will append data:

• No ads and tracking
• In-depth guides for developers and sysadmins at Opensourceflare✨
• Join my Patreon to support independent content creators and start reading latest guides:
  • How to set up Redis sentinel cluster on Ubuntu or Debian Linux
  • How To Set Up SSH Keys With YubiKey as two-factor authentication (U2F/FIDO2)
  • How to set up Mariadb Galera cluster on Ubuntu or Debian Linux
  • A podman tutorial for beginners – part I (run Linux containers without Docker and in daemonless mode)
  • How to protect Linux against rogue USB devices using USBGuard

Join Patreon ➔

Verify it:
cat /tmp/data.txt

How to save the output of a command to a file in bash using tee command

The tee command read from standard input and write to standard output and files. The syntax is as follows for writing data into the file:
command | tee file.txt
Want to append data? Try
command | tee -a output.txt

Examples

Display output of the date command on screen and save to the file named /tmp/output.txt. If the output.txt already exists, it gets overwritten:
$ date | tee /tmp/output.txt
$ cat /tmp/output.txt
Same as above but append to the given files, do not overwrite file:
$ pwd | tee -a /tmp/test.txt
$ echo «Today is $(date)» | tee -a /tmp/test.txt
$ hostnamectl | tee -a /tmp/test.txt
$ cat /tmp/test.txt

The above commands will append the output to the end of the file, just like the shell >> operator as explained earlier.

I/O redirection summary for bash and POSIX shell

Shell operator	Description	Overwrite existing file?
command > output.txt	Save terminal output (standard output) to a file named output.txt	Yes
command >> output.txt	Append terminal output (standard output) to a file named output.txt	No
command	Takes standard input from output.txt file	N/A
command 0	Takes standard input from output.txt file	N/A
command 1> output.txt	Puts standard output to output.txt file	Yes
command 1>> output.txt	Appends standard output to output.txt	No
command 2> output.txt	Puts standard error to output.txt	Yes
command 2>> output.txt	Appends standard error to output.txt file	No
command &> output.txt	Puts both standard error and output to output.txt	Yes
command > output.txt 2>&1	<POSIX> Puts both standard error and output to file named output.txt	Yes
command &>> output.txt	Appends both standard error and output to file named output.txt	No
command >> output.txt 2>&1	<POSIX> Appends both standard error and output to file called output.txt	No
command | tee output.txt	Puts standard output to output.txt while displaying output on screen	Yes
command | tee -a output.txt	Appends standard output to output.txt while displaying output on screen	No
command |& tee output.txt	Puts both standard output and error to output.txt while displaying output on terminal	Yes
command 2>&1 | tee output.txt	<POSIX> Puts both standard output and error to file named output.txt while displaying output on terminal	Yes
command |& tee -a output.txt	Append both standard output and error to file called output.txt while displaying output on terminal	No
command 2>&1 | tee -a output.txt	<POSIX> Append both standard output and error to file named output.txt while displaying output on terminal	No

Conclusion

You learned how to write the output to the file in Linux or Unix-like system when using bash or POSIX shell. We have:

1. /dev/stdin (standard input) — File descriptor 0 is duplicated.
2. /dev/stdout (standard output) — File descriptor 1 is duplicated.
3. /dev/stderr (standard error) — File descriptor 2 is duplicated.

See I/O redirection documentation for more information. We can read bash man page as follows using the man command:
man bash

🐧 Get the latest tutorials on Linux, Open Source & DevOps via

Источник