Has anyone had a case where connecting to the ESX HyperVisor with VMWare Infrastructure Client would time out, despite the valid credentials being given? The message I received was: response from HOSTNAME timed out.
On the box, two of three VMWare Images had stopped responding to pings, and inspecting the logs showed that apache 2.2 httpd had segfaulted all it's child processes (it was running Drupal 6) the other image was running MySql, which again wasn't responding. Only a reverse proxy image was still running and responding to pings and http requests (503s).
Being unable to connect to the VMW ESX console, we ended up having to hard reboot the machine, which when it came up had the 2 images that had not been responding in a stopped state that should have started automatically but didn't.
All of this was very frustrating, and actually leads to two questions, first does this mean something is wrong with ESX and how could it be prevented? second why would httpd have segfaulted (is it because of something that occurred in ESX?) and the separate mysql machine have gone down as well (or did that cause the http segfault)?
I have had a couple of random issues like this with my 3.5 server. A call to support corrected this issue.
From what I remember increase the amount of memory that the service console can use. Max it out. I think the max setting is like 800 Megs or something. There's also something which they can disable within the web server. Sorry I can't remember any more details than that.
Try logging onto the ESX itself rather than connecting to VM Ware Center. That sometimes works for me...