Snapshotting a running VM on iscsi leads to I/O error => fixed in SVN, will be fixed in 3.1.4

[bauer40]

I am running a most current debian VM on top of VBox 3.1.2 on Solaris 10 U6u7, kernel 141415-04. This VM uses an ISCSI disk provided by the same physical host.

When trying to snapshot the running machine by selecting "Take snapshot" from the "Machine" menu and trying to take a snapshot, the progress starts, takes a few seconds and aborts again.

After that, further disk I/O requests of the guest lead to write errors.

The most recent guest additions are installed.

Set the priority to minor due to the not so frequent setup using ISCSI.

[bauer40]

VBox.log

[bauer40]

XML configuration file of the VM

[Klaus Espenlaub]

Confirmed. Caused by an incorrect empty string check. I didn't try it, but maybe you could work around this by explicitly specifying LUN 0. That's the immediate problem which prevents the VM from continuing.

[bauer40]

Replying to klaus:

Confirmed. Caused by an incorrect empty string check. I didn't try it, but maybe you could work around this by explicitly specifying LUN 0. That's the immediate problem which prevents the VM from continuing.

Good to know. I suppose you mean I should temporary use the --lun option :

VBoxManage addiscsidisk --server some.host.name --target iqn.... --lun 0

Will this be fixed?

[Klaus Espenlaub]

Sure it will be fixed.