Controlling and managing processes

Jobs

CTRL + z            # send current process into the backgroud
jobs                # view all jobs running in bg
fg                  # open most recent job
fg 1                # open job 1
<program> &         # start <program> in the bg

ps

View processes currently running on your system:

Process ID (PID) differentiates each process from the next on your server
- Server only knows PID

Field	Description
`PID`	Process ID
`TTY`	The TTY that the process is tied to
`STAT`	Status code - the current state of the process. D (sleep), Z (defunct), T (stopped), S (interruptible sleep), R (run queue)
`TIME`	Total amount of CPU time consumed by this process
`COMMAND`	Command being run by the process

Interruptible sleep means that the process is waiting for input and cannot handle new signals
defunct is a zombie, waiting for the parent process to clean it up
TTY is teletypewriter - a terminal that accepts input and manages the output
- Historically, when users connected to large mainframe computers with a keyboard and monitor
- Now, its our terminal that is connected to a machine locally or with SSH
- Most processes run on a TTY
- pts/0 is a virtual (pseudo) terminal, and pts is a designation assigned to a terminal emulator

ps                                  # list of processes run by user
pidof <name>                        # get PID of named process
ps a                                # list processes for all users
ps au                               # list processes with usernames
ps aux                              # list all processes for all users with usernames, and 
                                    # include daemons or bg processes, not limited to TTY processes
ps aux | grep <keyword>             # search processes for <keyword> - commonly used
ps u -C <keyword>                   # show processes that have <keyword>
ps aux --sort=-pcpu                 # sort processes using most CPU
ps aux --sort=-pcpu | head -5       # show top 5 processes using most CPU since boot
ps aux --sort=-pmem | head -5       # show top 5 processes using most memory since boot

Change process priority

You can change the priority of running processes:

Not very common today because of RAM available in modern machines
- Some data processing and deep learning processes might require process priority changes
nice values range from -20 to 19
nice sets priority of processes when you start them
The higher the nice value, the lower the priority
renice changes priority of running processes
You can’t change the niceness of a process that you do not own
Requires sudo to lower the nicencess of a process

Think of the nice (NI) value as how nice a process is to other processes on the system, so the higher the niceness value, the nicer the process and the lower the priority.
lower the niceness value, the nicer the process and the and higher the priority. So, 10 is a higher priority than a process running NI of 20.

nice

Set the priority of a new process:

nice -n <num> <program>

renice

Renice changes priority of running processes.

By default, processes begin with a nice value of 80. If you use renice with the -n 10 argument, then you are adding 10 to the default NI value, resulting in 90. If you again run renice -n 5 -p <PID>, then you are lowering the current NI value by 5, resulting in 85.

ps -l                               # view priority and nice cols
renice -n <num> -p <PID>
renice -n <num> <process-name>


# --- Example --- #
# 1. View current PRI and NI
ps -l
F S   UID     PID    PPID  C PRI  NI ADDR SZ WCHAN  TTY          TIME CMD
0 S  1000    1093    1091  0  80   0 -  2254 do_wai pts/0    00:00:03 bash
0 T  1000   12047    1093  0  80   0 -  8981 do_sig pts/0    00:00:00 vim
0 R  1000   12196    1093 99  80   0 -  2729 -      pts/0    00:00:00 ps

# 2. Change priority with renice
renice -n 10 -p 12047
12047 (process ID) old priority 0, new priority 10

# 3. Verify the new PRI NI values
ps -l
F S   UID     PID    PPID  C PRI  NI ADDR SZ WCHAN  TTY          TIME CMD
0 S  1000    1093    1091  0  80   0 -  2254 do_wai pts/0    00:00:03 bash
0 T  1000   12047    1093  0  90  10 -  8981 do_sig pts/0    00:00:00 vim
0 R  1000   12205    1093 99  80   0 -  2729 -      pts/0    00:00:00 ps

Misbehaving processes

Use the kill command to send a process a SIGTERM (15) signal to stop gracefully:

a signal is sent to a process by the kernel, a process, or manually
- signals tell the process about a request or a change
- learn more at man 7 signal
SIGKILL (9) force closes a process
- Can have harmful effects, but probably not likely
- Use case: zombie process. Might even need a reboot bc not assigned a CPU.

kill <PID>              # send SIGTERM
killall <process>       # send SIGTERM to all processes that match <process>

kill -9 <PID>           # send SIGKILL
killall -9 <process>    # not recommended

Managing system processes

Daemons run processes in the background and typically start at boot:

When we install an app to run as a service, the system confiugures it to start at boot
Services are managed by the init system, or PID 1 - systemd
“service”, “daemon”, and “unit” are synonymous
Manage units with systemctl
Check vendor preset in Loaded line of systemctl status <unit> to see if unit starts automatically after installation

systemctl

systemctl                       # lists all available units
systemctl list-units            # same as above
systemctl status <unit>         # check status of service, enabled, recent logs

systemctl start <unit>          # starts the unit
systemctl stop <unit>           # stops the unit
systemctl restart <unit>        # restarts the unit (stops then starts)
systemctl reload <unit>         # activate new configurations without bringing unit down (ex: Apache)
                                # NOTE: not all units have 'reload' option

systemctl enable <unit>         # tell unit to start when server boots
systemctl disable <unit>        # tell unit not to start when server boots
systemctl enable --now ssh      # enable a unit and also start it

Scheduling with cron

cron can run a process at a specific time:

Each user has their own cron configurations, called a crontab
crontab is a list of cron jobs, one per line
cron jobs require fully-qualified commands, such as /usr/bin/apt
- find path with which
root can schedule system-wide tasks

crontab -l                      # view your crontab
sudo crontab -u <user> -l       # view <user>'s crontab

crontab -e                      # edit crontab (create new cronjob)
EDITOR=vim crontab -e           # set default editor for crontab actions

# --- Example crontab entries --- #
3 0 * * 4 /usr/local/bin/backup.sh      # run cleanup.sh at 12:03 AM each Thursday
* 0 * * * /usr/bin/apt update           # run apt update at midnight each night