User Fred Weston had a error when he tried converting his amazon ec2 windows 2008 boxes to Vmware workstation vms. He experienced a BSOD that pointed to storport.sys
You need to somehow get access to the target disks. You can use VMware Workstation to mount the target VMDK or you can temporarily attach the disk with the Windows directory to another Windows VM (in this case you should open Disk Manager and select Rescan Disks command from the Actions menu and the new virtual disk from the converted Amazon VM should become visible). After that: * run regedit on that machine * select HKLM hive * from the File menu select 'Load Hive' command * navigate to the target SYSTEM registry hive (it is under WINDOWS\system32\config\system *on the target disk* that was attached to the VM, but *not* the currently running Windows directory) * pick a name for the new key (e. g. ec2-system) * navigate to ec2-system\ControlSet001\Services * search for the three drivers and delete the whole keys * then select again the ec2-system key under HKLM * from File menu select 'Unload Hive' command and close Regedit * unmount the mounted VMDK or detach the target virtual disk from the "temporary" VM
Advantages - easy to get past the 2 TB disk limit. This was more of an issue in the past as you can get around this in ESXi 5.1 with a physical RDM - required for official support with MS clustering - (see the support protocols section) - http://kb.vmware.com/kb/1037959
Disadvantages - the backend storage network is exposed to a guest OS - portability of the VM may be hindered - i.e. you have to make sure the destination host will allow for the VM to have all the network access it needs. That'll probably be the case, but it's more difficult to verify than if the storage is presented via ESXi. - a snapshot of the VM also requires another snapshot of the iSCSI LUN that the guest OS accesses - you need to configure multipathing with the guest OS
Received this error while rescanninghis storage adaptors:
Error while rescanning adapter vmhba36. Error was Unable to scan VMkernel SCSI subsystem for new devices.Sysinfo error on operation returned status : Maximum allowed SCSI paths have already been claimed.
Problem:
His count of the scsi paths was no where near 1024, which is the toal number of paths allowed.
He had 12 paths
Example of where to look in the GUI:
Answer he found:
The gui didn’t count dead paths but they are counted against him:
Used esxcfg-mpath -lto find the list of paths from the command line
Over 1000 dead paths? that's a lot of paths.
Rebooting fixed him up.
Could have also done this:use esxcfg-rescan -d adaptor
In this case adaptor would be vmhba32 which has the dead path
here is esxcfg-mpath -l getting an error in a state of a path
rescan the hba
Now if we look for any dead paths we don’t find one.