We have a Microsoft Windows 2003 Cluster connecting to g a number of LUNS from one of our NetApps. Recently we had to forcebly shut down a cluster node which resulted in the remaining node being unable to connect to the disks dropped by the deceased node. During this time we saw the following error for over 20 minutes on our FAS (ONTAP 7.3) until I took action to reset the connection to the LUN.
Fri Aug 19 10:32:21 BST [HOSTNAME: iscsi.notice:notice]: ISCSI: Initiator (iqn.1991-05.com.microsoft:HOSTNAME) sent LUN Reset request, aborting all SCSI commands on lun 2
Does anyone have any idea why this LUN lock could not be cleared? The second node was off and thus not able to renew its reservation of the LUN. It just looks like the NetApp has it's knickers in a twist over something...
But assuming the machines are switched off and nothing sends any LUN reset commands when does the LUN get released so that other initiators can connect to it. I ask because we turned both machines off for 10 minutes and nobody could connect to the LUN as it was locked ...