ceph-fuse: handle suspend and resume more gracefully
Right now, suspend and resume are treated as "mds disconnect" events and we basically lose everything or risk data loss. It would be great if we could, instead:
1) detect suspend is happening and clean up most of the client,
2) reconnect to the cluster on resume,
3) and try and be transparent about it to applications with open files and things.
See the discussion at http://comments.gmane.org/gmane.comp.file-systems.ceph.devel/19594, https://wiki.archlinux.org/index.php/pm-utils#Advanced_configuration and http://www.freedesktop.org/software/systemd/man/systemd-sleep.html might be useful.