When I worked as a professional sys-admin, I sincerely didn't understand why sys-admins were paid so well.
I remember thinking that anyone could do what I'm doing, and I was surprised at how I knew programmers making less than I did.
Today, having hired devops and re-training myself to do the work, I realize why sys-admins/devops are paid well.
Firstly, it's a niche industry. While there are many programmers available, there are fewer people who understand the principles of high quality system administration.
Secondly, most people who are trained in this are already working or in high demand. Demand drives pay.
Thirdly, it's a changing field that moves- in some ways- faster than software.
It's easy to find someone who think they know devops because they run their own Linux laptop, but someone who really knows both the tools and the methodology of system administration is actually quite rare.
Here’s one of the many reasons why I stick to simple, replicable, and quickly restorable setups. When you don't have control over your data or rely solely on a specific service, everything could go wrong in an instant.
Taking a minute to appreciate how damn resilient my services/servers in the house are at the moment. They're all deployed slipshod and manually, too. Despite some annoyances, we've come a long way with software.
The mail server on which the power supply died had been rebooting spontaneously and hanging on startup a couple of times a day, and I spent a lot of time sifting through the logs to see what might be triggering it. I finally scripted a bunch of startup scripts to send an alert, circumvent the hang, restart the mail services, and just let it do its reboot thing.
With the new PS it seems to be stable. Was it a bad power supply all along? We'll never know now...
Is there any text-based, ideally distributed monitoring software out there? I want a TUI that shows me (e.g. with green/red highlights) the reachability of hosts (simple ping checks) while I'm doing network maintenance. Like a really simple, curses based nagios? It would probably look like a bloomberg terminal.. (I know this wouldn't be too hard to implement.. but I can't really believe it's not already out there...) #linux#TUI#monitoring#network#networking#sysadmin#software#curses
I'll even make Rat Pancakes to encourage them to attend.
Backers of all levels get four books, the release party, and the (unfortunate) video of me doing the Happy Dance. Plus whatever stretch goals get unlocked in the remaining days.
Microsoft’s new Copilot+ feature will record everything you are doing on your computer for some reason, but it will only work on new Arm hardware for now. Plus Apple’s weird iOS bug that restored deleted files and photos, and sharing files over the Internet from a NAS on your LAN.
Hmm I probably have the most ridiculous #robotstxt for a #Misskey instance right now lol. I just want to let #Mojeek and #Marginalia crawl #Makai and make sure to keep out #Google and the AI scrapers... :satrithink:
Just encountered the #EDIFACT standard while in touch with our logistics firm. Why the hell is this nothing readable and yet another useless standard that could've been implemented in #protobuf#json or something modern....
And the best part is everyone uses a subset or has some customizarions which forces us to implement it again and again
#sysadmin#cybersécuritay
Un type se connecte en IMAP sur TLS au serveur, depuis une machine en Chine, et se déconnecte avec l'alerte TLS "bad certificate".
C'est un certificat CAcert. S'il ne te plait pas, tu dégages !
#sysadmin
Un démon plantait à peu près tous les jours. Je mets en place un script qui teste s'il est toujours là et le démarre sinon, et, depuis, le démon ne plante plus (là, trois jours d'uptime). #Alpine