Hi, I'm trying to troubleshoot some errors on some hard drives.
I have a Supermicro SYS-6028R-TRT server that I have put x8 10TB hard drives in, specifically 8 WD100EZAZ, which were shucked (from WD My Book 10TB).
They are connected straight through the servers backplane into the SATA ports on the motherboard. They don't go to any LSI card, the only LSI card in this server is just to connect the JBOD (This was a High Availability ZFS system from Nextron which is now being rebuilt for other purposes)
When I put the disks in and began setting up FreeNAS, I ran badblocks command on every drive to test them:
badblocks -v -b 4096 -ws /dev/adaX
They all returned corruption errors
Pass completed, 941095 bad blocks found. (0/0/941095 errors)
So my first thought was that it couldn't be a problem with the drives, since all drives are brand new and ALL return errors, it has to be something like the LSI card, HBA, or something else. Or an old LSI firmware, or that it's running in IR firmware instead of IT. But apparently the LSI card is not used by them as I stated above and it's running IT.
So I emailed Supermicro and they said
I have another SYS-6028R-TRT that is identical (They were 2 nodes connected to a JBOD), I'm gonna try to move my FreeNAS usb and hard drives to this server and try to run badblocks again.
But I need advice, what do you think it could be? Could it be something with badblocks, the software itself? The SAN worked perfectly before, we have never had problems with the backplane until now. Other commands I should try?
I have a Supermicro SYS-6028R-TRT server that I have put x8 10TB hard drives in, specifically 8 WD100EZAZ, which were shucked (from WD My Book 10TB).
They are connected straight through the servers backplane into the SATA ports on the motherboard. They don't go to any LSI card, the only LSI card in this server is just to connect the JBOD (This was a High Availability ZFS system from Nextron which is now being rebuilt for other purposes)
When I put the disks in and began setting up FreeNAS, I ran badblocks command on every drive to test them:
badblocks -v -b 4096 -ws /dev/adaX
They all returned corruption errors
Pass completed, 941095 bad blocks found. (0/0/941095 errors)
So my first thought was that it couldn't be a problem with the drives, since all drives are brand new and ALL return errors, it has to be something like the LSI card, HBA, or something else. Or an old LSI firmware, or that it's running in IR firmware instead of IT. But apparently the LSI card is not used by them as I stated above and it's running IT.
So I emailed Supermicro and they said
and I also asked if it could be the integrated SATA controller, or if its possible to upgrade/view its firmwareIt is very unlikely that all eight slots on the backplane are faulty. I suggest installing a known working hard drive into each slot and verifying its functionality. You can also try swapping the cable from the motherboard to the backplane as well.
So I tried other drives, I found some 1TB drives I had laying around, and they did not get any badblock errors.. Maybe I need to try bigger drives? I didn't specify any block size on this badblock command when I tried the 1TB drives.The backplane in the system is a passthrough backplane and does not have any firmware controlling the backplane or hard drives.
I have another SYS-6028R-TRT that is identical (They were 2 nodes connected to a JBOD), I'm gonna try to move my FreeNAS usb and hard drives to this server and try to run badblocks again.
But I need advice, what do you think it could be? Could it be something with badblocks, the software itself? The SAN worked perfectly before, we have never had problems with the backplane until now. Other commands I should try?