6 ms·

System V init was decent enough for the time, but today it's relatively garbage compared to the alternatives. Services die? Oh well, I guess they'll just stay

System V init was decent enough for the time, but today it's relatively garbage compared to the alternatives. Services die? Oh well, I guess they'll just stay dead. At least systemd has a wrapper that restarts them. Also the "write a shell script with magical comments in it and lots of low-level bash" is not an elegant solution to any problem. Fedora's been pretty good about supporting legacy sysvinit style scripts and there will always be a way to use them.

brokenparser 13y ago
System V init could restart a service, that's how you got to log in after logging out. Even so, you actually need init to restart your important service process that should never die in the first place? Now that's just sloppy ;)
darkarmani 13y ago
> Services die? Oh well, I guess they'll just stay dead. At least systemd has a wrapper that restarts them. If services dies, there is a problem that is still going to exist after restarting.
- astrodust 13y ago
  I've had named blow out and it doesn't get restarted, it just stays dead. Rebooting the server is a heavy handed fix. systemd will kick it back into gear if it drops.
  - brokenparser 13y ago
    Use nsd and/or unbound, they're more robust and faster than bind.
    - astrodust 13y ago
      Good to know, hadn't really had a chance to explore those yet but will. No software is entirely bug-free, so I do like having them relaunch on failure rather than stay dead.
      - brokenparser 13y ago
        Before systemd, Nagios was employed to do that and alert you when it happens. It's still a good idea to have proper monitoring set up, be it Nagios/Icinga, Zabbix or something you know will fit the bill. In good fashion of Unix philosophy, a combination of components works really well. This repository has most of them: https://github.com/monitoringsucks/tool-repos https://github.com/monitoringsucks/tool-repos

brokenparser 13y ago

System V init could restart a service, that's how you got to log in after logging out. Even so, you actually need init to restart your important service process that should never die in the first place? Now that's just sloppy ;)

darkarmani 13y ago

> Services die? Oh well, I guess they'll just stay dead. At least systemd has a wrapper that restarts them. If services dies, there is a problem that is still going to exist after restarting.

astrodust 13y ago
I've had named blow out and it doesn't get restarted, it just stays dead. Rebooting the server is a heavy handed fix. systemd will kick it back into gear if it drops.
- brokenparser 13y ago
  Use nsd and/or unbound, they're more robust and faster than bind.
  - astrodust 13y ago
    Good to know, hadn't really had a chance to explore those yet but will. No software is entirely bug-free, so I do like having them relaunch on failure rather than stay dead.
    - brokenparser 13y ago
      Before systemd, Nagios was employed to do that and alert you when it happens. It's still a good idea to have proper monitoring set up, be it Nagios/Icinga, Zabbix or something you know will fit the bill. In good fashion of Unix philosophy, a combination of components works really well. This repository has most of them: https://github.com/monitoringsucks/tool-repos https://github.com/monitoringsucks/tool-repos