It may be storage issue on that particular host....all the VMs are running from shared storage, but this host boots from a local storage.
I can't access any of the logs files... cat /var/log/<any file> locks the SSH session forever.
In dmesg (this one works)... there is indeed something that might be related (it's filtered):
2013-08-22T12:05:37.622Z cpu9:12332924)<4>hpsa 0000:05:00.0: Abort request on C6:B0:T0:L1
2013-08-22T12:07:52.195Z cpu13:2211)<4>hpsa 0000:05:00.0: Abort request on C6:B0:T0:L1
2013-08-22T12:07:55.426Z cpu0:16799858)<4>hpsa 0000:05:00.0: Abort request on C6:B0:T0:L1
2013-08-22T12:25:37.707Z cpu13:16808864)<4>hpsa 0000:05:00.0: Abort request on C6:B0:T0:L1
Those errors started to appear less than an hour before vcenter lost this host.
For the "df", I wish not to try it... I cannot open new SSH session because VMware refuse to allocate PTYs... too many sessions are frozen.