On Tue, Jun 21, 2016 at 12:15:14AM +0200, Christoph Anton Mitterer wrote: > On Mon, 2016-06-20 at 23:35 +0200, Alexander Wirt wrote: > > On Mon, 20 Jun 2016, David Kalnischkies wrote: > > I guess the nagios check shouldn't use 'apt-get upgrade', but that > > > depends on what it is supposed to show (aka what its users expect) > > > and > > > what it actually uses (based on "critical updates" I guess it is > > > using > > > its own code, perhaps a binding…) but both I don't know hence > > > reassigning. > > I don't think so. The plugin is expected to run on all distributions > > having > > apt-get, apt is fairly new. > > I don't think David implied it should use apt(8), did he?
He meant that he has no idea what should be used rather than whatever is used at the moment as he doesn't even know what that is. He isn't even sure its a bug in nagios, but it surely isn't one in apt where he would have closed it if the mail hadn't implied that reassign would be an option. That said, I am NEVER advocating to use 'apt' in scripts. The command itself will complain about being used in scripts… > Also @David: Or is there any other, even more Or is there any other, > Or is there any other, even more portable/reliable/ancient-versions- > compatible way to determine from apt how many updates (normal/security) APT doesn't make a difference between "normal" and "security" because there isn't really one. Updates in stable aren't 'just for fun' anyhow and any detection based on which repository the update comes from is bound to fail as security updates eventually end up being distributed over non-security sources, too. > would be available & installed (when packages would also be > added/removed to do that - of course, I don't think check_apt should The obvious answer is "dist-upgrade" as it is the only command which is allowed to install new/remove packages. But as said, I have no idea where and how information is extracted from and what users expect to be told by it (I don't even know if the display is in text, as graph or as a dance-performance by a group of monkeys) so I am really not the person to be asked about that stuff. Best regards David Kalnischkies
signature.asc
Description: PGP signature
