[Pkg-fcoe-general] Bug#796609: fcoe-utils: Has init script in runlevel S but no matching service file

[Felipe Sateler]

On 25 January 2017 at 14:36, Jacob Luna Lundberg <jacob at gnifty.net> wrote:
>
> Hi Felipe,
>
> On Wed, Jan 25, 2017 at 11:50:33AM -0300, Felipe Sateler wrote:
>> Apparently, fcow-utils as currently packaged is not usable in
>> debian.
>
> It is plenty usable, just not if you are using systemd.  Yes, I am aware
> policy at this point requires systemd support.

OK, sorry. The impression I had from your earlier message is that it
didn't work[1].

>> Jacob Luna Lundberg, Liang Guo, could you please review and upload
>> this patch?
>
> I can't review the systemd components.  I do not use systemd and I am
> unfamiliar with its configuration.  Also, I am not a DD, so although I
> could upload the patch to the repository, I can't upload a package.

I can sponsor your upload, but as I have never used fcoe, I do not
want to upload such a large NMU without maintainer input.

> There are (as mentioned) a few bugs in the current package, so if there
> has been an upstream release, we should really incorporate that as well.
> Some of the patches merged into this jumbo patch also sound like a good
> idea for inclusion into Debian.  I can (will) look into all of the
> non-systemd parts in a month or so when my schedule settles a bit.  I
> would be inclined to just accept the systemd parts as submitted given
> they seem to have seen some testing already.

Seems like a good plan. I can provide input on the systemd-specific parts:

0030-fcoe.service-Add-foreground-to-prevent-fcoemon-to-be.patch

=> Seems to me it is better to change the unit to Type=forking
instead? This way, systemd would know when the monitor is ready.

0038-Makes-fcoe.service-starting-before-network.patch

=> this appears to have a slight malformation (a / is missing in the
header, this it edits etc/systemdfcoe.service)

debian/rules:

=> it seems weird to me that the socket is not started by default.
=> I would process the socket before the daemon, so that the daemon
cannot be started by the socket when removing the package.


If you have any questions, I'm always available to help.

[1] https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=796609#22

-- 

Saludos,
Felipe Sateler