Troubleshooting

Health & Performance

CPU

# cat /proc/cpuinfo # lscpu # sensors <- necesita "apt install lm-sensors" # htop # mpstat -P ALL 1 # ps aux --sort=-%cpu | head -10 <- cei mai mari 10 consumatori de CPU

RAM

# free -h # awk '$3=="kB"{$2=$2/1024;$3="MB"} 1' /proc/meminfo | column -t <- conversie in MB # awk '$3=="kB"{$2=$2/1024**2;$3="GB";} 1' /proc/meminfo | column -t <- conversie in GB # ps aux --sort=-%mem | head -n 10 <- cei mai mari 10 consumatori de RAM

Disk

# df -h # df -hi <- arata si inode # du -hx --max-depth=1 /etc/ | sort -h <- afiseaza toate dir si fisierele din /etc/ sortate in ordinea crescatoare a dimensiunii # du -ah / | sort -rh | head -10 <- arata cele mai mari 10 fisiere sau directoare # watch -d -n 60 'df -H /dev/sd[a-z][0-9] | sort -r -k 5 -i' <- watch pe "df -h" pentru toate disk-urile montate # lsblk -f # fdisk -l

Status

# vmstat 1 # uptime

Services & Logs

# systemctl restart <serviciu> # systemctl start <serviciu> # systemctl stop <serviciu>

Journal

# journalctl -xb <- logurile de la ultimul boot # journalctl --since "2026-12-01 10:00" <- loguri incepand cu "2026-12-01 10:00" # journalctl -u <serviciu> -f <- loguri specifice unui serviciu # journalctl -xe <- erori recente din journal

Kernel

# dmesg -T # dmesg --level=err <- filtrare dupa erori

Network

# netstat -tulnp # ss -tulnp <- afiseaza porturile si socket cu LISTEN activ # ping -c 3 google.com <- ping catre IP/Domeniu # traceroute google.co <- traseul pachetelor catre IP/Domeniu # lsof -i :443 <- afiseaza porturile si conexiunile (exemplu pt 443) # tcpdump -i ens18 <- captura de pachete pe din interfata ens18 # mtr google.com <- combina ping si traceroute in timp real

Procese

# ps aux # ps aux | grep <serviciu> # ps axo stat,ppid,pid,comm | grep -w defunct <- identifica procese zombie # ps -eo pid,ppid,state,cmd | grep " D " <- afiseaza procese aflate in "Uninterruptible Sleep"