Systemd
systemd
journalctl
/etc/systemd/journald.conf
journalctl -u SERVICE- previous boots dmesg log
journalctl -k -b -1
- coredumpctl
sysctl
net.ipv4.icmp_echo_ignore_all = 1 in /etc/sysctl.conf or /etc/sysctl.d/90-override.conf to disable icmp
systemctl
- services
edit doesn't require daemon-reload- hooks within services like
ExecStartPost/ExecStopPost
- list
systemctl list-unit-files
- list failed
systemctl list-units --state failed
- deps
systemctl list-dependencies --reverse
- mask with
systemctl mask swap.target
- remote management with ssh
systemctl --host=user@somesuchhost.com status SERVICE.unit
- restart interval with
RestartSec=5
- cron like with a
service.timer containing a schedule like OnCalendar=Sunday *-*-* 4:00:00 and a service with the Type=oneshot
timedatectl
- systemd-timesyncd
timedatectl set-ntp true
set-local-rtc to set hardware clock to localtime instead of UTC
NetworkManager
/usr/lib/NetworkManager/conf.d/20-connectivity.conf pings out
- conflicts with
systemd-networkd
- configure for dnsmasq with
[main] dns=dnsmasq
resolvectl
- dns cache
- issue with long running process? Several weeks of uptime and lockup of dns request despite ping working…
resolvectl statistics and resolvectl status/etc/systemd/resolved.conf- enable
systemd-resolved and start
DNSStubListener listens on 127.0.0.53 127.0.0.1.54 for udp/tcp port 53
ln -sf ../run/systemd/resolve/stub-resolv.conf /etc/resolv.conf
- alternatively configure NetworkManager with
[main] dns=systemd-resolved
- for local dnsmasq set to no, add dnsmasq/localhost to DNS servers
- may require
bind-interfaces for dnsmasq config (binds 127.0.0.1 port 53)
misc
systemd-analyze blame for boot timessystemd-analyze security my-service.unit for sec settingssystemctl --user status for services in $HOME/.config/systemd/user/systemctl set-default or isolate with runlevels/targets rescue, multi-user and graphical- rate limited restart will disable units with
Restart=on-failure without StartLimitIntervalSec=0