Re: what init systems do you use ?

From: multiplexd <multiplexd_at_gmx.com>
Date: Mon, 13 May 2019 20:13:29 +0100

All,

On Fri, May 03, 2019 at 02:53:21AM +0200, Jeff wrote:
> what init systems do this list's subscribers use ?

I saw Guillermo's reply up-thread, so I thought I'd add my own two cents. My
main workstation at present is a Debian 9 box booted with s6 and s6-rc, with
system setup and teardown (i.e. stages 1 and 3) handled by some execline scripts
heavily adapted from those generated by s6-linux-init-maker 0.4.x.x.

It's been nearly two years since I set this up, so I don't recall all the
details, but I started out with a sysvinit Debian 9 system, and
reverse-engineered the boot sequence, which included a by-hand conversion of all
the necessary /etc/init.d scripts into s6-rc service definitions.

The boot process follows the general flow to be expected from a system using
s6-svscan as PID 1. The stage 1 script sets up an initial service directory,
spawns the script which starts stage 2 and then exec's into s6-svscan. Most of
the runtime state configuration and management is then handled through s6-rc in
stage 2. Stage 3 involves deactivating all the s6-rc managed services and then
signalling the PID 1 s6-svscan, which execs into a script which kills the
remaining processes, unmounts filesystems and then performs a halt, poweroff, or
reboot as necessary.

The particulars of the configuration I'm using are generally not portable to
other machines though, as it was a proof-of-concept I came up with a year or two
or go for a different system, which I then ported to my current workstation,
which was set up in a very similar manner. I've also had to make it largely
signal-compatible with sysvinit, as apt(8) sometimes attempts to reload the init
configuration upon package changes. The signal used to request a configuration
reload from sysvinit is SIGUSR1 if I recall correctly, however the default
behaviour of the version of s6-linux-init-maker I was using at the time was to
power off the machine upon receiving SIGUSR1 in PID 1, so I had to change the
signal configuration for s6-svscan.

That being said, it's been very stable and has largely stayed out my way, so
it's been worth the initial setup effort in the long run.

Cheers,
--multiplexd.
Received on Mon May 13 2019 - 19:13:29 UTC