I recently installed ESXi 5.0.0 and built a single VM on it. I then turned off the vm, enable ssh and connected through ssh, downloaded patches to the datastore1, went into maintenance mode, and applied the update. The update completed successfully. Old build number was 504890 and the update showed build number 515841. (I also tried this with the newest release build number 768111). I rebooted after the update completed successfully. As the host was restarting it came up with a purple screen with "PF Exception 14 in world 2659" error. At the bottom of the error it showed no place on disk to dump data. Debugger waiting (world 2659) - - no poer for remote debugger. "Escape" for local debugger. At this point I would either type reboot or do a hard restart and then the host would come back and be responsive again. The problem is that it is still showing BUILD Version 504890. So the update was not successful. Any comments would be appreciated. Thanks!
the reason it is showing build 504890 is because it went back to the old version after your updated version crashed.
so that is by design and is a good thing.
if you had to manully recover from a bad update you would have to hold shift+r
http://sparrowangelstechnology.blogspot.com/2012/07/esxi-recovering-from-bad-update-last.html
try updating again.
I believe that the issue is because I am running ESXi 5 with a free license. After going back through the User Agreements it seems that patches and upgrades are not permitted. Does this sound accurate?
i think you can patch via command line just not with update manager (GUI)
I too am receiving an error (purple screen) with an "#PF Exception 14 in world 2626" after I try to apply the patch "ESXi500-201207001.zip".
This week I took a test ESXi vSphere Hypervisor v4.1U1 (with Free License), and performed a forced migration using the Dell Custom ESXi vShpere v5.0U1. The current running build is 623860. I reapplied a Free License. The problem was I had the autostart bug.
I downloaded the patch to fix the autostart problem (http://blogs.vmware.com/vsphere/2012/07/vsphere-hypervisor-auto-start-bug-fixed.html) and stored it on my primary datastore (/vmfs/volumes/datastore1). Using SSH, I installed the patch via local command-line:
esxcli software vib install --depot=/vmfs/volumes/datastore1/ESXi500-201207001.zip
The appach applied successfully. However the next reboot while trying to load build number 768111, I received the exception 14. Of course it reverted back to build 623860 and loaded fine after a hard reset. Still unpatched and still experiencing the autostart bug.
While not exactly the same scenario bharp28, could be the same root cause. Any thoughts?
Thank you.
Hi bharp28 ,
Welcome to communites.
I am agree with sparrowangels .
We can overcome lot of small problem if will upgrade or install patches thorugh CLI.
I have tried multiple times with the same results. Still not sure what the issue is.
This sounds very similar to what I have been experiencing. Unfortunately I have not found a fix for it yet.