have the following situation:
My windows virtual machine has lost all changes since 4.5. I'm not sure why. Maybe it has something to do with vcb backups because there are lots of -delta files. Because I have the snapshot delta files on vmfs LUN, I tried to recover a point in time with these delta snapshots. Snapshot manager didn't show any snapshots at all. So at first I created a fresh snapshot with snapshot manager. Then I manually edited .vmx file to point to certain delta files and removed all delta files newer than the point in time that I was trying to restore. Finally I deleted all snapshots with snapshot manager.
This caused vmware\[init] to crash (from /var/log/messages):
May 16 19:02:25 esx3 VMware\[init]: connect: No such file or directory.
After that the ESX host lost connection to virtualcenter. I can't connect via VIClient directly to ESX host either. The virtual machines are running, but I can't manage them. If I try running vmware-cmd -l, it just hangs.
I have tried to restart /etc/init.d/mgmt-vmware, vmware-vmkauthd, vmware-vpxa, webAccess, xinetd many times without help. I don't want to restart the esx host. So what should I do?
vpxa.log says this:
\[2007-05-16 20:47:52.441 'App' 4332464 error] \[VpxVmdbCnx] Authd error: 514 Error connecting to hostd-vmdb service instance.
\[2007-05-16 20:47:52.441 'App' 4332464 error] \[VpxVmdbCnx] Failed to connect to host :902. Check that authd is running correctly
Looks like you've tried all the fixes - how's the disk space look (vdf -h)? I've found that some patience is required when restarting services and processes (May com back on it's own). Can you view the vm's through esxtop? How does top look - any run-away of defunct processes? Hostd and vpxa really running after the restart (ps -auxwww | grep vmware-hostd and vpxa)?
We encountered the exact same problem this morning.
While trying to remove a VM from a host, this one got disconnected and we were unable to have it Back. We were also unable to connect directly to that host with the VI-Client.
Restarting the vpxa did not do anything. mgmt-vmware neither.
The host came back after restarting the vmkauthd service.
It turns out, for me anyways, that after spending hours with VMware support, we had a network problem ??? But we were having roughly 50% packet loss which apparently caused the heartbeat errors. go figure. The networking problem resided at the physical switch, and once it was reset, everything came back to normal.