Hi,
I have a RHEL 5.0 linux target that locks up and goes into a read-only filesystem mode quite frequently.
Our setup is that there is 1 QNAP NAS box serving iSCSI targets that a host running ESXi 5.0 is connected to. (sidebar question: we used to have multiple hosts connected to the same iSCSI target and found out it is a bad thing - is that correct? Should it be that only 1 host is connected to an iSCSI target at a time?)
Looking on the QNAP, I see all the disks are fine and that there are no issues with the iSCSI targets.
However, running "dmesg" on the ESXi host shows this:
And it goes on and on and on for the iSCSI targets the host was attached to.
Is that the cause for why only my RHEL 5.0 machines lock up? What does that error mean and how can fix it?
This seems to be related with the QNAP box. If you check the VMware HCL, (www.vmware.com/go/hcl) only one or two QNAP devices are supported for ESXi 5 - check if your is one of this ones before start testing or using it.
Anyway, the error you is seeing calls attention to the disks inside the box:
H:0x0 = Host Status = 0x0 = OK
D:0x2 = Device Status = 0x2 = Check Condition
P:0x0 = Plugin Status = 0x0 = OK
0xb = ABORTED COMMAND
0x24 = INVALID FIELD IN CDB
When the error is on D, it is device related (disk). It would be good to check with QNAP support about it.
Also, having many initiators to one target is good, not bad. It's how things work normally on shared storage environments.