I've searched for answers to this one and there are several; but none that quite match our case, and none of the suggestions seem to provide a resolution.
It appears that a driver was updated on the fibre connector that caused the vCenter server to crash. Once the driver was rolled back VMs are running, but at the moment, we still can't run the vSphere client.
The services VirtualCenter Server and VirtualCenter Management Webservices will not start (fail with errors 1067 &1068).
In the vxpd.log there are a couple of key entries:
[2012-09-17 12:55:31.226 02796 warning 'Libs'] VmdbCtxGet: Failed to get /vpx/datastore/#5402/_private/managedObjImpl/ (Null cross reference)
[2012-09-17 12:55:31.226 02796 error 'App'] Group /vpx/group/#s5/ references invalid datastore /vpx/group/#s5/datastore/#_VeeamBackup\UPE-BKUP01-UKPAR (1)/
This would indicate that the datastore is missing; however, it's still there and accessible. One link did suggest that this might be ODBC related, but it appears to be making a connection OK. Another link suggested that the VM had been moved to a new host with a different vCneter server; it's the same host as before, and we only have the one instance. Yet another link suggested that the DB had an invalid or duplicate entry; the entries are all the same as before and no duplicates.
Does anyone have any other suggestions that I could try out?