On Dec 31, 2020, at 13:10, Greg Woods <greg(a)gregandeva.net> wrote:
Yes. The format is "Got automount request for /pub, triggered
by #PID# (#command#): 1 Time(s)
I suppose I should provide more info. /pub is a place where I store a bunch of stuff. In
particular there are our photo and music collections, and rsync'ed Fedora repos (I
rsync F32 and F33 repos once a day, then I use the local repo to perform updates on my 6
Fedora systems).
The problem here is that, by the time I see one of these entries and investigate, the PID
is for a process that is long gone, and I'm not getting any useful info about what
command is triggering this.
As an example, here is a message that is the result of a known access (doing a "dnf
update"):
pub.automount: Got automount request for /pub, triggered by 253560 (dnf): 1 Time(s)
I suspect that dnf and other tools are looking at all the entries in / and seeing if they
are mountpoints, to determine if there is enough space when running rpm operations. That
way you don’t run out of space mid-transaction.
Because your automount is mounted in /, I’d expect this kind of thing to trigger. I
believe just running something like ‘ls -F /‘ ought to trigger the automount.
This is why I always hang my automounts off a subdirectory of /, such as /mnt or /net.
—
Jonathan Billings