42
Condor
6y

*deploys new VPS*
Click clack tap.. alright, done.
*notices that I accidentally made an Ubuntu 14.04*
Well shit... Guess I'll have to update that immediately to 18.04 then.
*logs in, immediately disables SSH password auth*
# systemctl restart sshd
> systemctl: command not found.
What the fuck..?
What was the command for that old init again.. >_<
# /etc/init.d/ssh restart

WHY THE FUCK IS THIS UBUNTU STILL USING THAT OLD INIT?!! Goddamit, Canonical living up to the philosophy of its Debian counterpart indeed!

Comments
  • 8
    Maybe because it was 1404?
    Have no idea what that init does
  • 5
    @mundo03 Of course, that's exactly the case. Heh, if there's anything good about this mistake, it'd be nostalgia. 14.04 was the first Linux system that I've ever used. Now I recall that at that time, the migration in Ubuntu to systemd was in full swing, haha. Who knew that I'd come back to this 4 years later. Oh well. Updating to 16.04 now, with another one to 18.04 pending before I go to sleep :)
  • 2
    i think upstart was an 15 or 16 thing. 14 had sysv.
  • 2
    @stop I think so yes.. sysvinit if memory serves me right. Eh, it's a good init system, but the days of custom init scripts are gone really. Too many shitadmins out there that would fuck them up, or make them a royal maintenance clusterfuck.
  • 2
    @Condor Yess this is my life. I got on a Centos 6 server the other day at work and had to restart Apache 2.2 (this sentence makes me cringe) “systemctl not found wtf? Oh yeah... /etc/init.d/httpd restart”
  • 3
    @Diactoros I just use the service command
  • 1
    @andros705 me too on debian.
  • 0
    @stop @andros705 "it's a feature" 🙃
    @Diactoros and it's never been considered to upgrade? 😰
    @xzvf that's why I like systemd so much. During startup it does a significant amount of parallelization. Especially on multicore systems it makes a really big difference. And I find it a lot easier to use to be honest.
  • 1
    And here I am, still thinking of systemctl as, "new."
  • 1
    @Condor Oh man, it has, that’s the worst part. The company has a legacy PHP app and they had one guy running the infrastructure without any server management background. The whole fleet is on CentOS 6, hundreds of clients, no configuration management, no deployment process beyond:
    $scp ./* root@legacy-CentOS-box

    I have my hands full.
  • 0
    @Diactoros ... Holy shit.. how do you even survive such a thing? 😰
Add Comment