There is normally 3 sources of failure when you can still ping the server.
1) Storage issues, this causes hostd to become "hung" and unresponsive to any other requests, especially when you only have intermittent failures instead of a clean cut failure. As already mentioned /var/log/vmkernel.log is the one to look at.
2) vpxa failures as this is the agent which is responsible for the communication to vCenter. /var/log/vpxa.log
3) hostd failures unrelated to storage, vpxa needs proper hostd responses to communicate to vCenter, so if this agent fails your host would also show the described symptoms. /var/log/hostd.log
