Gestion des processus Linux Cheat Sheet
Aperçu général
La gestion des processus Linux implique la surveillance, le contrôle et l'optimisation des processus d'exécution et des ressources du système. Ce guide complet couvre les outils de suivi des processus, le traitement des signaux, le contrôle de l'emploi, la gestion des ressources du système et les techniques d'optimisation des performances essentielles pour les administrateurs de système et les professionnels de l'informatique.
C'est pas vrai. Attention : les commandes de terminaison de processus comme
kill -9peuvent causer une perte de données. Essayez toujours la terminaison gracieuse d'abord et assurez les sauvegardes appropriées.
Consultation et surveillance des processus
Informations de base sur le processus
# Display running processes
ps
ps aux # All processes with detailed info
ps -ef # Full format listing
ps -u username # Processes for specific user
ps -p PID # Specific process by PID
# Process tree view
pstree
pstree -p # Show PIDs
pstree username # Processes for specific user
# Real-time process monitoring
top
htop # Enhanced interactive process viewer
Surveillance avancée des processus
# Detailed process information
ps auxf # Process tree format
ps -eo pid,ppid,cmd,pcpu,pmem # Custom columns
# Process by resource usage
ps aux --sort=-pcpu # Sort by CPU usage
ps aux --sort=-pmem # Sort by memory usage
# Long-running processes
ps -eo pid,etime,cmd # Show elapsed time
# Zombie processes
ps aux|grep -w Z
ps -eo stat,pid,cmd|grep -w Z
```_
### Surveillance du système en temps réel
```bash
# Interactive process monitor
top
# Top commands:
# k - kill process
# r - renice process
# M - sort by memory
# P - sort by CPU
# q - quit
# Enhanced process monitor
htop
# htop features:
# F9 - kill process
# F7/F8 - nice/renice
# F6 - sort options
# F10 - quit
# System activity reporter
sar -u 1 10 # CPU usage every 1 second, 10 times
sar -r 1 10 # Memory usage
sar -d 1 10 # Disk activity
```_
## Contrôle des processus et signaux
### Types de signaux
```bash
# Common signals:
# SIGTERM (15) - Graceful termination
# SIGKILL (9) - Force kill (cannot be caught)
# SIGHUP (1) - Hangup (reload config)
# SIGSTOP (19) - Stop process
# SIGCONT (18) - Continue stopped process
# SIGINT (2) - Interrupt (Ctrl+C)
Processus de meurtre
# Kill by PID
kill PID
kill -15 PID # Send SIGTERM (default)
kill -9 PID # Force kill with SIGKILL
kill -HUP PID # Send hangup signal
# Kill by process name
killall process_name
killall -9 process_name
# Kill by pattern
pkill pattern
pkill -f pattern # Match full command line
pkill -u username # Kill user's processes
# Interactive process killer
top # Press 'k' then enter PID
htop # Press F9 then select process
Priorité du processus et valeurs de Nice
# View process priorities
ps -eo pid,ni,cmd
# Start process with specific priority
nice -n 10 command # Lower priority (+10)
nice -n -5 command # Higher priority (-5)
# Change running process priority
renice 10 PID # Set nice value to 10
renice -5 -p PID # Set nice value to -5
renice 5 -u username # Renice all user processes
Contrôle de l'emploi
Contexte et emplois nouveaux
# Run command in background
command &
# List active jobs
jobs
jobs -l # Show PIDs
# Bring job to foreground
fg
fg %1 # Bring job 1 to foreground
# Send job to background
bg
bg %1 # Send job 1 to background
# Suspend current job
Ctrl+Z
# Resume suspended job
fg # Resume in foreground
bg # Resume in background
Processus persistants
# Run command immune to hangups
nohup command &
nohup command > output.log 2>&1 &
# Detach from terminal
disown %1 # Disown job 1
disown -a # Disown all jobs
# Screen sessions
screen # Start new session
screen -S name # Named session
screen -r # Reattach to session
screen -ls # List sessions
# Tmux sessions
tmux # Start new session
tmux new -s name # Named session
tmux attach -t name # Attach to session
tmux list-sessions # List sessions
Surveillance des ressources du système
CPU Surveillance
# CPU usage overview
top
htop
vmstat 1 # Virtual memory statistics
iostat 1 # I/O and CPU statistics
# Per-CPU statistics
mpstat # Multi-processor statistics
mpstat -P ALL 1 # All CPUs every second
# CPU load averages
uptime
w # Who is logged in and load
cat /proc/loadavg # Load average file
Surveillance de la mémoire
# Memory usage
free
free -h # Human readable
free -m # In megabytes
free -s 1 # Update every second
# Detailed memory information
cat /proc/meminfo
vmstat 1 # Virtual memory stats
# Memory usage by process
ps aux --sort=-pmem|head -10
top -o %MEM # Sort by memory in top
# Memory map of process
pmap PID
pmap -d PID # Detailed mapping
Surveillance des E/S sur disque
# Disk usage
df -h # Filesystem usage
du -sh /path # Directory usage
# I/O statistics
iostat
iostat -x 1 # Extended stats every second
iotop # Real-time I/O monitoring
# Disk activity by process
iotop -o # Only show active processes
pidstat -d 1 # Per-process I/O stats
Surveillance des réseaux
# Network connections
netstat -tuln # TCP/UDP listening ports
ss -tuln # Modern replacement for netstat
ss -p # Show process names
# Network activity
netstat -i # Interface statistics
iftop # Real-time network usage
nethogs # Network usage by process
# Network connections by process
lsof -i # Network files
lsof -i :80 # Processes using port 80
Analyse des processus et débogage
Information sur le processus
# Detailed process info
cat /proc/PID/status
cat /proc/PID/cmdline
cat /proc/PID/environ
# Process file descriptors
lsof -p PID
ls -la /proc/PID/fd/
# Process memory maps
cat /proc/PID/maps
pmap PID
# Process threads
ps -T -p PID
top -H -p PID
Appels système et débogage
# Trace system calls
strace command
strace -p PID # Attach to running process
strace -f command # Follow forks
# Library calls
ltrace command
ltrace -p PID
# Debug running process
gdb -p PID
Analyse des résultats
# Process CPU usage over time
pidstat 1 # Per-process statistics
pidstat -p PID 1 # Specific process
# Memory usage tracking
pidstat -r 1 # Memory statistics
pidstat -d 1 # Disk I/O statistics
# System performance overview
vmstat 1 10 # Virtual memory stats
sar -A # All system activity
Services système et Daemons
Gestion systémique des services
# Service status
systemctl status service_name
systemctl is-active service_name
systemctl is-enabled service_name
# Start/stop services
systemctl start service_name
systemctl stop service_name
systemctl restart service_name
systemctl reload service_name
# Enable/disable services
systemctl enable service_name
systemctl disable service_name
# List services
systemctl list-units --type=service
systemctl list-units --failed
Registres de service
# View service logs
journalctl -u service_name
journalctl -u service_name -f # Follow logs
journalctl -u service_name --since today
# System logs
journalctl
journalctl -f # Follow system logs
journalctl --since "1 hour ago"
journalctl -p err # Error messages only
Automatisation et calendrier des processus
Cron Emplois
# Edit crontab
crontab -e
crontab -l # List cron jobs
crontab -r # Remove all cron jobs
# System cron directories
ls /etc/cron.d/
ls /etc/cron.daily/
ls /etc/cron.hourly/
ls /etc/cron.weekly/
ls /etc/cron.monthly/
# Cron log
tail -f /var/log/cron
Emploi (Planning unique)
# Schedule one-time job
at 15:30
at now + 1 hour
at tomorrow
# List scheduled jobs
atq
at -l
# Remove scheduled job
atrm job_number
Résolution de problèmes communs
CPU élevé Utilisation
# Find CPU-intensive processes
top -o %CPU
ps aux --sort=-pcpu|head -10
# CPU usage by command
pidstat -u 1
# System load investigation
uptime
vmstat 1
sar -u 1 10
Utilisation haute mémoire
# Memory-intensive processes
ps aux --sort=-pmem|head -10
top -o %MEM
# Memory leaks detection
valgrind --leak-check=full command
# Out of memory investigation
dmesg|grep -i "killed process"
grep -i "out of memory" /var/log/messages
Processus de zombies
# Find zombie processes
ps aux|awk '$8 ~ /^Z/ \\\\{ print $2 \\\\}'
ps -eo stat,pid,cmd|grep -w Z
# Kill parent of zombie
ps -o ppid= -p zombie_pid
kill parent_pid
Processus sans réponse
# Check if process is hung
kill -0 PID # Test if process exists
strace -p PID # See what process is doing
# Force termination
kill -TERM PID # Try graceful first
sleep 5
kill -KILL PID # Force if necessary
Optimisation des performances
Limites de processus
# View current limits
ulimit -a
cat /proc/PID/limits
# Set limits
ulimit -n 4096 # File descriptors
ulimit -u 1024 # Max user processes
ulimit -m 1048576 # Memory limit (KB)
# System-wide limits
cat /etc/security/limits.conf
CPU Affinité
# View CPU affinity
taskset -p PID
# Set CPU affinity
taskset -p 0x1 PID # CPU 0 only
taskset -p 0x3 PID # CPU 0 and 1
taskset -c 0,1 PID # CPU 0 and 1
# Start process with affinity
taskset -c 0 command
Ressources
- [Guide de gestion du processus Linux] (LINK_4)
- [Documentation systémique] (LINK_4)
- Outils d'analyse du rendement
- [Process Monitoring Best Practices] (LINK_4)
*Cette feuille de triche fournit des commandes complètes de gestion de processus pour les systèmes Linux. Toujours comprendre l'impact des commandes de contrôle de processus avant de les utiliser dans les environnements de production. *