There are two servers: Openfiler(v. 2.99) that exports a 5TB disk and Proxmox(v. 3.4) that uses this disk by iscsi-protocol.
When using this disk I see in /var/log/syslog and dmesg-output the following message:
kernel: sd 7: 0: 0: 0: [sde] Very big device. Trying to use READ CAPACITY (16).
How can I fix this problem? Or does the system not pay attention to it?
Thanks.
It's harmless, as long as you don't see the message Using 0xffffffff as device size immediately following your Very big device message. The kernel logs Trying to use READ CAPACITY (16) before it knows whether it'll work or not. So, in order to figure out if you've got a problem, you need to look to the next messages in the logs. As long as you don't see the Using 0xffffffff message, you're good.
The Linux SCSI drivers have to support everything from the tiniest, crappiest USB thumb drive all the way up to your disk and beyond. There are some heuristics in the kernel that attempt to divine whether it should issue a 16-byte (that's the 16 in your message) READ CAPACITY command that might not be supported, or just send the 10-byte one. They factor in whether the LU supports SPC-3 or protection information, or if it supports 16-byte commands at all. But it looks like in your case, it's simply falling through to "try the 10-byte command first, then do the 16-byte command if needed."