On Wed, 2016-07-27 at 08:16 -0400, Josh Boyer wrote:
On Wed, Jul 27, 2016 at 5:43 AM, Ruben Kerkhof
<ruben(a)rubenkerkhof.co
m> wrote:
>
> Hi all,
>
> Debian and Ubuntu have a package called unattended-upgrades.
> We have yum-cron which does something similar.
>
> One difference though is that unattended-upgrade drops a script in
> /etc/kernel/postinst.d/unattended-upgrades, which does this:
>
> #!/bin/sh
> if [ -d /var/run ]; then
> touch /var/run/reboot-required
> fi
>
> Using Ansible, I can quickly see which servers need a reboot due to
> a
> kernel upgrade.
>
> I think this would be nice to have in Fedora as well, the only
> question is which package
> should provide it.
>
> We have /etc/kernel/postinst.d too, but this directory is currently
> unowned.
> So if I'd wanted to add this to some package, which one should it
> be
> and what should it depend on?
>
> Alternatively, I could create a new package, let's call it 'reboot-
> required'.
>
> Thoughts?
Why would you want this to be something packaged? We have 'reboot
recommended' in our bodhi update metadata, and that seems like a much
better place for it. Otherwise, you run into cases where multiple
packages want to write/own the file, etc.
Also, I think "recommended" is really the appropriate terminology
here. There is very little that _requires_ a reboot to be done after
it is installed.
How can this metadata be leveraged with automation? I have the dnf
tracer plugin which I believe is using this metadata to tell me when I
need to reboot, but what if I have this in a cron job?
--
John Florian <john.florian(a)dart.biz>