Hi all,
So, I've recently (in the past week or so) started having an issue
with _an_ SL6.4 system (only one), wherein sometimes it will fail to
NFS mount otherwise-available mounts (the NFS servers will not log an
error or success when this transpires), and IO operations to some
mountpoints will return EIO.

But the mounts themselves, insofar as I can quickly see, are perfectly
fine - the servers are reachable by both this and every other machine,
no errors are being logged client or server-side, and the EIO errors
vanish within 5 minutes of anyone observing them.

The NFS servers are running SL5, SL6, or Illumos derivatives; some are
in another datacenter entirely, some are on the same switch as the
server in question.

It's only ever this server which encounters this difficulty, and only
in the past 2 weeks.

LOGGING="verbose" in sysconfig/autofs doesn't produce any output; I've
tried changing it to debug in the hope that it actually produces
something.

Anyone encountered this and know where I should be looking for a trail
of bloody fingerprints?

Thanks,
- Rich