On Wed 02 Dec 2015 at 16:59:23 (-0600), Don Armstrong wrote: > On Wed, 02 Dec 2015, Jape Person wrote: > > It's occurred to me that, though I have occasionally seen service > > shutown issues with sysv-init, they were never as pervasive or > > repetitve as it has been since switching to systemd as the init > > system. > > This is generally because sysv-init tends to not pay attention to > whether a particular service has actually stopped. Many init scripts > just send an appropriate signal, hope for the best, and return control. > > If your goal is just to shutdown the system regardless of what is > actually going on, that's great... but if you value your data, that's > not really a workable solution.
Which data do you mean? By the time I instigate a shutdown, I've saved my data and, if at home, copied to my server. So I'm hoping that's safe. By the time these Stop jobs start running, there are just system processes running. I don't really care about them as long as the filesystems get unmounted. Almost anything is better than a forced poweroff followed by fscking saying things at the next boot. > That said, most of these cases are bugs, not really cases where the > daemon is actually doing something; reporting the bugs when you run into > them will help them get fixed. At the time I had the timeout-increasing problem, I was still focussed on getting Start jobs to succeed, particularly when the kernel "forgot" to load binfmt_misc (for help on which, I thank you again). Cheers, David.