On Fri, Dec 27, 2019 at 12:32:27PM +0000, Laurent Bercot wrote:
> Is there real pressure to have this?

AFAIK, the only pressure is from systemd fanboys.  But this is indeed
a biggest criticism from them; we would be able to save quite a lot of
flamewars if the feature was simply there.  Nevertheless I understand
the feature will be, frankly, a vase.

> The problem with such a "babysitter" is that it would need to forward
> signals, much like execline's trap program. It's ugly, and I'd rather
> have people not do that any more than strictly necessary.

We will also need to handle disgusting PID files for double-forking
services.  And in order to be safe in case the service crashes before
the PID file is created, we will perhaps need some kind of startup
deadline.  A big can of worms.

> As for cgroups-related chainloaders, I could probably write some in
> s6-linux-utils, but wasn't the cgroups interface designed to make
> sure those operations are trivial to implement as small scripts?

Well, this is a good idea.  I can even provide such library scripts in
slew, but the libraries will be `rc'-specific, and not used in the
traditional (exec()-based) sense of chainloading.

> I don't, for several reasons, one of which is that Google's homemade
> supervisor (which is... not great) is called "babysitter", and it
> triggers cringey memories.


My current OpenPGP key:
RSA4096/0x227E8CAAB7AA186C (expires: 2020.10.19)
7077 7781 B859 5166 AE07 0286 227E 8CAA B7AA 186C

Reply via email to