February 12, 201115 yr So I just upgraded from 4.6 to 4.7 and when I booted up I got the error saying "Two or more disks are wrong." I attached a screenshot and syslog. Not really an expert when it comes to these issue and wanted to see what I should do next. Thanks in advance. syslog_4.7.txt
February 12, 201115 yr easy just got to your disks page unassign both wrong drives and assign them right again... in other words just swap them on your drives page when you go back to the main page after it should be all OK and you will be able to start the array i better shut up looking second time at the pic i saw they were at the right spot ... HPA issue as BRiT said
February 12, 201115 yr No, not so easy. You definitely have HPA on those two drives. You need to address the HPA issue before they will function in unRAID 4.7 or later. Jan 2 20:23:08 Tower kernel: ata3.00: [color=red]HPA detected[/color]: current 1953523055, native 1953525168 Jan 2 20:23:08 Tower kernel: ata3.00: ATA-7: [color=red]Hitachi HDS721010KLA330[/color], GKAOAB0A, max UDMA/133 Jan 2 20:23:08 Tower kernel: ata4.00: [color=red]HPA detected[/color]: current 1953523055, native 1953525168 Jan 2 20:23:08 Tower kernel: ata4.00: ATA-7: [color=red]Hitachi HDS721010KLA330[/color], GKAOAB0A, max UDMA/133 See the HPA link in my signature for further details.
February 12, 201115 yr Author I tried following the instructions and could not remove the HPA. Here is what I got: Tower login: root Linux 2.6.32.9-unRAID. root@Tower:~# hdparm -N p1953525168 /dev/sde /dev/sde: setting max visible sectors to 1953525168 (permanent) SET_MAX_ADDRESS failed: Input/output error max sectors = 1953523055/7368112(1953525168?), HPA setting seems invalid (bug gy kernel device driver?) root@Tower:~# hdparm -N p1953525168 /dev/sdd /dev/sdd: setting max visible sectors to 1953525168 (permanent) SET_MAX_ADDRESS failed: Input/output error max sectors = 1953523055/7368112(1953525168?), HPA setting seems invalid (bug gy kernel device driver?) root@Tower:~# And attached is the syslog Can I click on the "I'm sure I want to do this" and start the array? The new size it shows is larger then the old, so not sure if it would let me do that. Or will that make matters worse? Can I go back to 4.6 for now? I have a M4A785-M motherboard on the way and should be here next week. I also picked up some 2TB drives and will be replacing the disk 2 and disk 3 that have HPA. I am getting away from Gigabyte boards due to the HPA issues and this is not the first time. syslog.txt
February 12, 201115 yr You can probably set a new initial configuration and then have unRAID recalculate parity based on the new configuration or revert to the 4.6 version you were running.
February 12, 201115 yr Author I ended up just reverting to 4.6 for now and it is back to working. Once I get my new motherboard I will replace disks 2 and 3 with 2TB drives and hopefully since the board is Asus a HPA is not created. Before I install the new drives into my server, I plan on running the preclear scrpit (which will be the first time for me). Should I use the -A option since I will not be running 4.7 yet. The drives are the HITACHI Deskstar 5K3000 2TB.
February 12, 201115 yr I ended up just reverting to 4.6 for now and it is back to working. Once I get my new motherboard I will replace disks 2 and 3 with 2TB drives and hopefully since the board is Asus a HPA is not created. Before I install the new drives into my server, I plan on running the preclear scrpit (which will be the first time for me). Should I use the -A option since I will not be running 4.7 yet. The drives are the HITACHI Deskstar 5K3000 2TB. If you plan on using the new disks on 4.6 then do not use the -A option. (the disks would not be recognized as pre-cleared by 4.6 and it would clear and partition them on its own.) You can use the "-A" option if you do not plan on assigning the disks until after you migrate to 4.7. No disk partitioned for MBR-4k-aligned will ever be recognized by 4.6 as being an unRAID disk. Joe L.
Archived
This topic is now archived and is closed to further replies.