| 1 |
On Wed, Dec 21, 2016 at 8:36 AM, Corbin Bird <corbinbird@×××××××.net> wrote: |
| 2 |
> |
| 3 |
> The old manual method of configuration is extremely flexible, you can |
| 4 |
> get the "who-knows-where-it-came-from-component" to work. The new |
| 5 |
> "automagic" of udev / systemd .... forget it. At least with script based |
| 6 |
> init systems I could change the run level to fix Xorg problems. |
| 7 |
|
| 8 |
udev and systemd operate based on text configuration files that are |
| 9 |
declarative in nature. |
| 10 |
|
| 11 |
You can certainly change the "run level" in systemd (what you'd call a |
| 12 |
runlevel in openrc would be a target in systemd). You can even pass |
| 13 |
the default target on the kernel command line, or change the default |
| 14 |
target in /etc. As with openrc they aren't numbered and you aren't |
| 15 |
limited to any particular number of them. There are some standard |
| 16 |
ones out of the box, like multi-user, emergency, getty, basic, etc. |
| 17 |
|
| 18 |
> |
| 19 |
> The systemd configuration files are designed for programmers, not |
| 20 |
> technicians. And their is a HUGE difference between "programmers" and |
| 21 |
> "technicians". Different aptitudes, different skills. The old .conf |
| 22 |
> files, technicians can easily handle. Requiring everyone to be a |
| 23 |
> programmer is a really bad idea. |
| 24 |
> |
| 25 |
|
| 26 |
The only "configuration" files openrc supports for services are shell |
| 27 |
scripts, as opposed to declarative configuration files used by |
| 28 |
systemd. Now, openrc init.d shell scripts might source configuration |
| 29 |
from some text file in /etc/conf.d, but there is nothing that prevents |
| 30 |
systemd units from doing the same. On Gentoo we stick the settings in |
| 31 |
drop-in files instead, but these are no more complex. |
| 32 |
|
| 33 |
Here is an example of a Gentoo systemd drop-in: |
| 34 |
/etc/systemd/system/ntpdate.service.d/00gentoo.conf |
| 35 |
[Service] |
| 36 |
Environment="SERVER=0.gentoo.pool.ntp.org 1.gentoo.pool.ntp.org |
| 37 |
2.gentoo.pool.ntp.org 3.gentoo.pool.ntp.org" |
| 38 |
|
| 39 |
|
| 40 |
That hardly requires programming to understand. |
| 41 |
|
| 42 |
|
| 43 |
And here is the entire ntpdate unit file: |
| 44 |
/usr/lib/systemd/system/ntpdate.service |
| 45 |
[Unit] |
| 46 |
Description=Set time via NTP using ntpdate |
| 47 |
After=network-online.target nss-lookup.target |
| 48 |
Before=time-sync.target |
| 49 |
Wants=time-sync.target |
| 50 |
Conflicts=systemd-timesyncd.service |
| 51 |
|
| 52 |
[Service] |
| 53 |
Type=oneshot |
| 54 |
ExecStart=/usr/sbin/ntpdate -b -u $SERVER |
| 55 |
RemainAfterExit=yes |
| 56 |
|
| 57 |
[Install] |
| 58 |
WantedBy=multi-user.target |
| 59 |
|
| 60 |
|
| 61 |
No programming there either. |
| 62 |
|
| 63 |
Most of the stuff that is hard to understand in the file are the |
| 64 |
dependencies, and that is just because you need to learn the |
| 65 |
terminology that systemd uses, though most of that is straightforward. |
| 66 |
The After= line is roughly equivalent to "use net dns" in openrc, |
| 67 |
though systemd has a lot more virtuals defined out of the box and |
| 68 |
they're more granular. For example, systemd distinguishes between an |
| 69 |
interface existing, and an interface having an IP/etc, while on openrc |
| 70 |
we have just one virtual that covers the latter. |
| 71 |
|
| 72 |
I know, it almost sounds like the systemd design is intended to |
| 73 |
support running a diverse service ecosystem. Go figure... |
| 74 |
|
| 75 |
-- |
| 76 |
Rich |
Archives