On Mar 16, 2015 12:46 PM, "Robert Elz" <kre%munnari.oz.au@localhost> wrote:
>
> Date: Mon, 16 Mar 2015 16:43:57 +0100
> From: Manuel Bouyer <bouyer%antioche.eu.org@localhost>
> Message-ID: <20150316154357.GA9193%asim.lip6.fr@localhost>
>
> | /proc is a pseudo filesystem, I'm not sure the iAvail number is
> | meaningfull.
>
> I think iAvail + iUsed == kern.maxproc so if iAvail == 0,
> iUsed == kern.maxproc and the system is running as many processes
> as permitted.
>
> kre
>
Yeah, what df reports for inodes for proc appears to be related to this sysctl:
# sysctl -e kern.maxproc
kern.maxproc=532
It was strange as I've never seen the original failure happen before, at least not in relation to proc. The other filesystems were just fine at the time. Nevertheless, I still failed to acertain the complete story behind my experienced behaviour. :^/
More testing is required. Sorry for the incomplete initial post.