Explain Shell pour comprendre une commande avant de la lancer.
C’était déjà bien pratique à l’époque des copier-coller depuis Stack Overflow.
Ça l’est sans doute encore plus aujourd’hui pour comprendre ce que vos IA vous demandent de taper dans le terminal
👉 explainshell.com/
C’était déjà bien pratique à l’époque des copier-coller depuis Stack Overflow.
Ça l’est sans doute encore plus aujourd’hui pour comprendre ce que vos IA vous demandent de taper dans le terminal
👉 explainshell.com/
November 10, 2026 at 7:30 PM
Explain Shell pour comprendre une commande avant de la lancer.
C’était déjà bien pratique à l’époque des copier-coller depuis Stack Overflow.
Ça l’est sans doute encore plus aujourd’hui pour comprendre ce que vos IA vous demandent de taper dans le terminal
👉 explainshell.com/
C’était déjà bien pratique à l’époque des copier-coller depuis Stack Overflow.
Ça l’est sans doute encore plus aujourd’hui pour comprendre ce que vos IA vous demandent de taper dans le terminal
👉 explainshell.com/
Laatste reden/tijd voor reboot soms bekeken?
last -x | grep -E reboot
journalctl -b1 | grep Reboot
('t Was toch geen UI herstart hé? Ik draai alle console dinges in tmux/screen om onafhankelijk van UI te werken. Geleerd in de tijd dat desktop nog kwikkel kwakkel was in xfree86, voorganger van X :)
last -x | grep -E reboot
journalctl -b1 | grep Reboot
('t Was toch geen UI herstart hé? Ik draai alle console dinges in tmux/screen om onafhankelijk van UI te werken. Geleerd in de tijd dat desktop nog kwikkel kwakkel was in xfree86, voorganger van X :)
November 13, 2026 at 8:00 PM
Laatste reden/tijd voor reboot soms bekeken?
last -x | grep -E reboot
journalctl -b1 | grep Reboot
('t Was toch geen UI herstart hé? Ik draai alle console dinges in tmux/screen om onafhankelijk van UI te werken. Geleerd in de tijd dat desktop nog kwikkel kwakkel was in xfree86, voorganger van X :)
last -x | grep -E reboot
journalctl -b1 | grep Reboot
('t Was toch geen UI herstart hé? Ik draai alle console dinges in tmux/screen om onafhankelijk van UI te werken. Geleerd in de tijd dat desktop nog kwikkel kwakkel was in xfree86, voorganger van X :)
Linux commands are easy to memorize. Knowing when to use them is what matters.
ps checks processes.
ss reveals network connections.
journalctl examines system logs.
chmod manages permissions.
For SOC investigations, these commands help turn system activity into useful evidence.
ps checks processes.
ss reveals network connections.
journalctl examines system logs.
chmod manages permissions.
For SOC investigations, these commands help turn system activity into useful evidence.
November 13, 2026 at 5:58 AM
Linux commands are easy to memorize. Knowing when to use them is what matters.
ps checks processes.
ss reveals network connections.
journalctl examines system logs.
chmod manages permissions.
For SOC investigations, these commands help turn system activity into useful evidence.
ps checks processes.
ss reveals network connections.
journalctl examines system logs.
chmod manages permissions.
For SOC investigations, these commands help turn system activity into useful evidence.
by the way, this block of output is so long that I had to learn a new `journalctl` option just to print all of it at once. 270+ KB of pure text
November 11, 2026 at 3:30 AM
by the way, this block of output is so long that I had to learn a new `journalctl` option just to print all of it at once. 270+ KB of pure text
Diagnose with `journalctl -u sshd | grep "Failed password"`. Fix: Deploy fail2ban with `bantime=1h`, `findtime=10m`, `maxretry=3`. Verify via `fail2ban-client status sshd`. Save on setup costs via https://www.topcashback.it/ref/member9909468601?subid=blueskysocia_1115 (#ad).
---
**Further Technical
---
**Further Technical
November 15, 2026 at 2:56 PM
Diagnose with `journalctl -u sshd | grep "Failed password"`. Fix: Deploy fail2ban with `bantime=1h`, `findtime=10m`, `maxretry=3`. Verify via `fail2ban-client status sshd`. Save on setup costs via https://www.topcashback.it/ref/member9909468601?subid=blueskysocia_1115 (#ad).
---
**Further Technical
---
**Further Technical
journald may keep up to 10% of the filesystem by default, capped at 4G. On a small homelab SSD, that adds up.
Check: journalctl --disk-usage
Trim: journalctl --vacuum-size=500M
Cap: drop a .conf into /etc/systemd/journald.conf.d/
[Journal]
SystemMaxUse=500M
Vacuum only deletes archived files.
Check: journalctl --disk-usage
Trim: journalctl --vacuum-size=500M
Cap: drop a .conf into /etc/systemd/journald.conf.d/
[Journal]
SystemMaxUse=500M
Vacuum only deletes archived files.
November 10, 2026 at 5:24 PM
journald may keep up to 10% of the filesystem by default, capped at 4G. On a small homelab SSD, that adds up.
Check: journalctl --disk-usage
Trim: journalctl --vacuum-size=500M
Cap: drop a .conf into /etc/systemd/journald.conf.d/
[Journal]
SystemMaxUse=500M
Vacuum only deletes archived files.
Check: journalctl --disk-usage
Trim: journalctl --vacuum-size=500M
Cap: drop a .conf into /etc/systemd/journald.conf.d/
[Journal]
SystemMaxUse=500M
Vacuum only deletes archived files.
Diagnose `sshd` brute-force attempts via `journalctl -u ssh`. Deploy fail2ban with iptables backend and aggressive rate-limiting. For resilient, managed security infrastructure, deploy on GreenGeeks: https://www.greengeeks.com/track/affiliatechproject/cp-ffltc/lp-1?subid=blueskysocia_1113 (#ad).
--
--
November 13, 2026 at 12:14 PM
Diagnose `sshd` brute-force attempts via `journalctl -u ssh`. Deploy fail2ban with iptables backend and aggressive rate-limiting. For resilient, managed security infrastructure, deploy on GreenGeeks: https://www.greengeeks.com/track/affiliatechproject/cp-ffltc/lp-1?subid=blueskysocia_1113 (#ad).
--
--
"Troubleshooting Linux in the Cloud: 4 Classic Commands That Save the Day on AWS EC2" by Kelvin Ferreira
#linux #amazon-ec2 #systems-manager #compute #beginner-projects
#linux #amazon-ec2 #systems-manager #compute #beginner-projects
Troubleshooting Linux in the Cloud: 4 Classic Commands That Save the Day on AWS EC2
Learn how four classic Linux commands (lsblk, ss, journalctl, and namei) help you quickly diagnose storage, network, service, and permission issues on AWS EC2.
community.aws
November 12, 2026 at 1:00 PM
"Troubleshooting Linux in the Cloud: 4 Classic Commands That Save the Day on AWS EC2" by Kelvin Ferreira
#linux #amazon-ec2 #systems-manager #compute #beginner-projects
#linux #amazon-ec2 #systems-manager #compute #beginner-projects
294 lines of Python that read a crontab and write systemd .service and .timer units, so your jobs land in journalctl. Intermediate level. #python #linux #systemd https://www.valtersit.com/python/cron-to-systemd-timer-converter/
Cron to Systemd Timer Converter
www.valtersit.com
November 10, 2026 at 5:40 AM
294 lines of Python that read a crontab and write systemd .service and .timer units, so your jobs land in journalctl. Intermediate level. #python #linux #systemd https://www.valtersit.com/python/cron-to-systemd-timer-converter/
"Job for X.service failed because the control process exited with error code" only tells you that something failed.
Where to find why, how to test the config before a restart, and what 203/EXEC means.
ohsudo.com/articles/job...
Where to find why, how to test the config before a restart, and what 203/EXEC means.
ohsudo.com/articles/job...
"Job for nginx.service failed because the control process exited with error code": how to find the cause
Fix "Job for X.service failed because the control process exited with error code": read systemctl status and journalctl, test the config, and fix the usual causes.
ohsudo.com
November 13, 2026 at 2:04 PM
"Job for X.service failed because the control process exited with error code" only tells you that something failed.
Where to find why, how to test the config before a restart, and what 203/EXEC means.
ohsudo.com/articles/job...
Where to find why, how to test the config before a restart, and what 203/EXEC means.
ohsudo.com/articles/job...
That shim deadlock on 2.2.2 was infuriating because the daemon process looks healthy while tasks silently freeze on SIGTERM. Having an LLM parse the raw journalctl trace saves an hour of staring at unhelpful containerd logs.
November 7, 2026 at 4:39 PM
That shim deadlock on 2.2.2 was infuriating because the daemon process looks healthy while tasks silently freeze on SIGTERM. Having an LLM parse the raw journalctl trace saves an hour of staring at unhelpful containerd logs.
A slow desktop right after an update is usually one bad package, not a broken install. journalctl --list-boots and apt's history log tell you which boot and which batch.
https://www.fosslinux.com/163673/linux-mint-slow-after-update.htm
#FOSSLinux #Linux
https://www.fosslinux.com/163673/linux-mint-slow-after-update.htm
#FOSSLinux #Linux
November 9, 2026 at 12:09 AM
A slow desktop right after an update is usually one bad package, not a broken install. journalctl --list-boots and apt's history log tell you which boot and which batch.
https://www.fosslinux.com/163673/linux-mint-slow-after-update.htm
#FOSSLinux #Linux
https://www.fosslinux.com/163673/linux-mint-slow-after-update.htm
#FOSSLinux #Linux
Been there—put the server in a dedicated systemd unit (separate user + working directory, Restart=on-failure) and use `journalctl -u <unit>` to follow logs. A scheduled, tested world backup makes experiments much less stressful.
November 8, 2026 at 10:10 AM
Been there—put the server in a dedicated systemd unit (separate user + working directory, Restart=on-failure) and use `journalctl -u <unit>` to follow logs. A scheduled, tested world backup makes experiments much less stressful.
O Journalctl vai te dizer certinho o que tá rolando. Mas eu chuto que seja alocação mesmo.
November 6, 2026 at 6:17 PM
O Journalctl vai te dizer certinho o que tá rolando. Mas eu chuto que seja alocação mesmo.
Your cron job died with exit 137 and its log just stops. That's SIGKILL, and SIGKILL doesn't say who sent it. How to prove it was the OOM killer (and when it wasn't), why the job can't log its own death, and how to cap one job's memory: #cron #linux #sysadmin #selfhosted
Cron job killed with exit 137 or signal 9: was it the OOM killer?
Exit 137 means SIGKILL, and SIGKILL doesn't say who sent it. How to tell the OOM killer from a timeout or a person with journalctl -k, memory.events and Docker's OOMKilled flag, why the job can't log its own death, and how to cap one job's memory so it can't take the box down.
runwisp.com
November 7, 2026 at 6:27 PM
Your cron job died with exit 137 and its log just stops. That's SIGKILL, and SIGKILL doesn't say who sent it. How to prove it was the OOM killer (and when it wasn't), why the job can't log its own death, and how to cap one job's memory: #cron #linux #sysadmin #selfhosted
bought a vertical stand for the secondary monitor three years ago and it still just displays `journalctl -u docker -f` in a 9pt font while I stare at the left screen like an idiot.
November 6, 2026 at 12:15 PM
bought a vertical stand for the secondary monitor three years ago and it still just displays `journalctl -u docker -f` in a 9pt font while I stare at the left screen like an idiot.
🚨 EUVD-2024-34378
📊 5.5/10
🏢 Red Hat
📝 A flaw was found in the FreeIPA API audit, where it sends the whole FreeIPA command line to journalctl. As a consequence, during the FreeIPA installation ...
🔗 https://euvd.enisa.europa.eu/vulnerability/EUVD-2024-34378
#cybersecurity #infosec #cve #euvd
📊 5.5/10
🏢 Red Hat
📝 A flaw was found in the FreeIPA API audit, where it sends the whole FreeIPA command line to journalctl. As a consequence, during the FreeIPA installation ...
🔗 https://euvd.enisa.europa.eu/vulnerability/EUVD-2024-34378
#cybersecurity #infosec #cve #euvd
November 5, 2026 at 1:03 PM
🚨 EUVD-2024-34378
📊 5.5/10
🏢 Red Hat
📝 A flaw was found in the FreeIPA API audit, where it sends the whole FreeIPA command line to journalctl. As a consequence, during the FreeIPA installation ...
🔗 https://euvd.enisa.europa.eu/vulnerability/EUVD-2024-34378
#cybersecurity #infosec #cve #euvd
📊 5.5/10
🏢 Red Hat
📝 A flaw was found in the FreeIPA API audit, where it sends the whole FreeIPA command line to journalctl. As a consequence, during the FreeIPA installation ...
🔗 https://euvd.enisa.europa.eu/vulnerability/EUVD-2024-34378
#cybersecurity #infosec #cve #euvd
I haven't tried it on Fedora, but Fedora Magazine's hibernation guide warns: with Secure Boot on and the kernel in lockdown mode, hibernation does not work without signed hibernation images. The note dates from 2022, so I'd check journalctl -b for "Lockdown" lines first.
November 2, 2026 at 2:25 PM
I haven't tried it on Fedora, but Fedora Magazine's hibernation guide warns: with Secure Boot on and the kernel in lockdown mode, hibernation does not work without signed hibernation images. The note dates from 2022, so I'd check journalctl -b for "Lockdown" lines first.
#DietPi JournalCTL logs are not persistent by default.
You can enable persistent logs they get put in /var/log, but DietPi has as tmpfs because of RAMLog. DOH.
Creating systemd-journal-flush.d/override.conf and a var-log-journal.mount to store in /var/lib/journal, seems like it does the trick!
You can enable persistent logs they get put in /var/log, but DietPi has as tmpfs because of RAMLog. DOH.
Creating systemd-journal-flush.d/override.conf and a var-log-journal.mount to store in /var/lib/journal, seems like it does the trick!
November 3, 2026 at 7:49 PM
#DietPi JournalCTL logs are not persistent by default.
You can enable persistent logs they get put in /var/log, but DietPi has as tmpfs because of RAMLog. DOH.
Creating systemd-journal-flush.d/override.conf and a var-log-journal.mount to store in /var/lib/journal, seems like it does the trick!
You can enable persistent logs they get put in /var/log, but DietPi has as tmpfs because of RAMLog. DOH.
Creating systemd-journal-flush.d/override.conf and a var-log-journal.mount to store in /var/lib/journal, seems like it does the trick!
Intermediate Python script, 283 lines, ships systemd journal to Loki so journalctl stops being your only search tool. Runs per unit with custom labels. #monitoring #python #systemd https://www.valtersit.com/python/systemd-journal-to-loki-forwarder/
Systemd Journal to Loki Forwarder
www.valtersit.com
November 3, 2026 at 3:30 AM
Intermediate Python script, 283 lines, ships systemd journal to Loki so journalctl stops being your only search tool. Runs per unit with custom labels. #monitoring #python #systemd https://www.valtersit.com/python/systemd-journal-to-loki-forwarder/
New ebook: Linux Without the Guesswork. The command line, permissions, systemd services and timers, journalctl, SSH keys, disks and a fixed troubleshooting order, for homelabbers who copy-paste commands.
Free sample: https://payhip.com/b/YLEK5
Book: https://payhip.com/b/LV4qz
Free sample: https://payhip.com/b/YLEK5
Book: https://payhip.com/b/LV4qz
November 2, 2026 at 2:50 PM
New ebook: Linux Without the Guesswork. The command line, permissions, systemd services and timers, journalctl, SSH keys, disks and a fixed troubleshooting order, for homelabbers who copy-paste commands.
Free sample: https://payhip.com/b/YLEK5
Book: https://payhip.com/b/LV4qz
Free sample: https://payhip.com/b/YLEK5
Book: https://payhip.com/b/LV4qz
Stop brute force: Check `journalctl -u sshd | grep "Invalid user"`. Fix: `fail2ban-client add sshd iptables-multiport` with `banaction = iptables-multiport`. Ensure `maxretry=3`. For robust SSH security, deploy on Hosting.com (#ad): https://hosting.com/?utm_source=fp_affiliate&utm_medium=15437677&ut
November 1, 2026 at 1:36 PM
Stop brute force: Check `journalctl -u sshd | grep "Invalid user"`. Fix: `fail2ban-client add sshd iptables-multiport` with `banaction = iptables-multiport`. Ensure `maxretry=3`. For robust SSH security, deploy on Hosting.com (#ad): https://hosting.com/?utm_source=fp_affiliate&utm_medium=15437677&ut
Hot take: for homelab jobs, systemd timers beat cron.
Persistent=true catches up on runs missed while the box was off, logs land in journalctl, and systemctl list-timers shows what fires next.
Cron still wins at "one line, done". Which camp are you in?
Persistent=true catches up on runs missed while the box was off, logs land in journalctl, and systemctl list-timers shows what fires next.
Cron still wins at "one line, done". Which camp are you in?
![Dark terminal-style card showing a systemd timer unit file /etc/systemd/system/backup.timer: [Unit] Description=Nightly backup; [Timer] OnCalendar=*-*-* 03:00:00, a comment 'catch up on runs missed while powered off', Persistent=true, RandomizedDelaySec=10m; [Install] WantedBy=timers.target; then the command systemctl enable --now backup.timer.](https://cdn.bsky.app/img/feed_thumbnail/plain/did:plc:lu3yyb7kchssv5oovqgoraly/bafkreifw35xyar4ba74kxiiplkluujbi5gtklcaruqhyaroiqg6yqhso5y)
October 29, 2026 at 5:40 PM
Hot take: for homelab jobs, systemd timers beat cron.
Persistent=true catches up on runs missed while the box was off, logs land in journalctl, and systemctl list-timers shows what fires next.
Cron still wins at "one line, done". Which camp are you in?
Persistent=true catches up on runs missed while the box was off, logs land in journalctl, and systemctl list-timers shows what fires next.
Cron still wins at "one line, done". Which camp are you in?