[systemd-devel] why does nofail imply no After= in /etc/fstab

[Tom Gundersen]

On Thu, Jan 16, 2014 at 2:01 AM, Dominique Michel
<dominique.michel at vtxnet.ch> wrote:
> Le Wed, 15 Jan 2014 20:20:18 +0100,
> Zbigniew Jędrzejewski-Szmek <zbyszek at in.waw.pl> a écrit :
>
>> I was a bit surprised that for mount points the dependency
>> Before=local-fs.target is only added when nofail is not used.
>> This seems to be a concious decision (added by Lennart in
>> 155da457, and then survived all the refactorings by Tom
>> and Thomas...). Do we still want this behaviour?

I do not recall a very strong reason for letting this survive. I
suppose x-systemd.automount should be sufficient for when we don't
want a mount to delay local-fs.target?

> If systemd goal is to provide a mean for the user or the administrator
> to implement policies, you must provide a way for the administrator to
> set this himself. If systemd goal is to provide a policy, you must
> decide which policy you want to implement.

The admin has the full power over how these things shall behave when
they write their own .mount files. What Zbigniew is asking is what the
default behavior should be when we interpret the legacy /etc/fstab
(which is less powerful than .mount files, so some choices have to be
made).

-t