Hey guys, I had a disk go bad on me and the replacement I had from a previous RMA was a WD 1.5T drive that I had been using in my i5, with a Gigabyte GA P55M-UD2 mobo, desktop for a bit. Well after a bit of maneuvering I was able to move everything off of the failed drive and ran initconfig and proceeded to preclear the 1.5T WD15EARS with the 4k alignment. I assigned it to the array and rebuilt parity and it seems to be working fine. This just happened a couple days ago, and yesterday i was looking at mymain and noticed it said there was a possible hpa on my parity drive. I have had a couple drives with hpa for a while though I know when i upgraded to 4.7 those were not there. They most likely were added into the array after the upgrade.

C2SEE-O-P, Celeron 420, 4gb ram, adaptec 1430SA,SUPERMICRO AOC-SASLP-MV8

 

I tried hdparm -N and it didnt seem to work, so I am wondering if this is a false positive because of the 4k alignment or if I should pull the disk and try to remove the hpa outside the array.

Also, I was going through my syslog which I admit is like trying to read heiroglyphics for me but I noticed a bunch of errors. This was the most disturbing block of errors..

 

Not sure why it says C2SEA there cause its clearly marked C2SEE on the board...

 

Apr  4 01:15:08 Tower kernel: Hardware name: C2SEA

Apr  4 01:15:08 Tower kernel: Modules linked in: md_mod xor i2c_i801 i2c_core ata_piix sata_sil r8169 sata_sil24 sata_mv mvsas libsas scst scsi_transport_sas (Drive related)

Apr  4 01:15:08 Tower kernel: Pid: 6932, comm: hdparm Not tainted 2.6.32.9-unRAID #8 (Errors)

Apr  4 01:15:08 Tower kernel: Call Trace: (Errors)

Apr  4 01:15:08 Tower kernel:  [<c102449e>] warn_slowpath_common+0x60/0x77 (Errors)

Apr  4 01:15:08 Tower kernel:  [<c10244c2>] warn_slowpath_null+0xd/0x10 (Errors)

Apr  4 01:15:08 Tower kernel:  [<c11b624d>] ata_qc_issue+0x10b/0x308 (Errors)

Apr  4 01:15:08 Tower kernel:  [<c11ba260>] ata_scsi_translate+0xd1/0xff (Errors)

Apr  4 01:15:08 Tower kernel:  [<c11a816c>] ? scsi_done+0x0/0xd (Errors)

Apr  4 01:15:08 Tower kernel:  [<c11a816c>] ? scsi_done+0x0/0xd (Errors)

Apr  4 01:15:08 Tower kernel:  [<c11baa40>] ata_sas_queuecmd+0x120/0x1d7 (Errors)

Apr  4 01:15:08 Tower kernel:  [<c11bc6df>] ? ata_scsi_pass_thru+0x0/0x21d (Errors)

Apr  4 01:15:08 Tower kernel:  [<f83ef69a>] sas_queuecommand+0x65/0x20d [libsas] (Errors)

Apr  4 01:15:08 Tower kernel:  [<c11a816c>] ? scsi_done+0x0/0xd (Errors)

Apr  4 01:15:08 Tower kernel:  [<c11a82c0>] scsi_dispatch_cmd+0x147/0x181 (Errors)

Apr  4 01:15:08 Tower kernel:  [<c11ace4d>] scsi_request_fn+0x351/0x376 (Errors)

Apr  4 01:15:08 Tower kernel:  [<c1126798>] __blk_run_queue+0x78/0x10c (Errors)

Apr  4 01:15:08 Tower kernel:  [<c1124446>] elv_insert+0x67/0x153 (Errors)

Apr  4 01:15:08 Tower kernel:  [<c11245b8>] __elv_add_request+0x86/0x8b (Errors)

Apr  4 01:15:08 Tower kernel:  [<c1129343>] blk_execute_rq_nowait+0x4f/0x73 (Errors)

Apr  4 01:15:08 Tower kernel:  [<c11293dc>] blk_eHPxecute_rq+0x75/0x91 (Errors)

Apr  4 01:15:08 Tower kernel:  [<c11292cc>] ? blk_end_sync_rq+0x0/0x28 (Errors)

Apr  4 01:15:08 Tower kernel:  [<c112636f>] ? get_request+0x204/0x28d (Errors)

Apr  4 01:15:08 Tower kernel:  [<c11269d6>] ? get_request_wait+0x2b/0xd9 (Errors)

Apr  4 01:15:08 Tower kernel:  [<c112c2bf>] sg_io+0x22d/0x30a (Errors)

Apr  4 01:15:08 Tower kernel:  [<c112c5a8>] scsi_cmd_ioctl+0x20c/0x3bc (Errors)

Apr  4 01:15:08 Tower kernel:  [<c11b3257>] sd_ioctl+0x6a/0x8c (Errors)

Apr  4 01:15:08 Tower kernel:  [<c112a420>] __blkdev_driver_ioctl+0x50/0x62 (Errors)

Apr  4 01:15:08 Tower kernel:  [<c112ad1c>] blkdev_ioctl+0x8b0/0x8dc (Errors)

Apr  4 01:15:08 Tower kernel:  [<c1131e2d>] ? kobject_get+0x12/0x17 (Errors)

Apr  4 01:15:08 Tower kernel:  [<c112b0f8>] ? get_disk+0x4a/0x61 (Errors)

Apr  4 01:15:08 Tower kernel:  [<c101b028>] ? kmap_atomic+0x14/0x16 (Errors)

Apr  4 01:15:08 Tower kernel:  [<c11334a5>] ? radix_tree_lookup_slot+0xd/0xf (Errors)

Apr  4 01:15:08 Tower kernel:  [<c104a179>] ? filemap_fault+0xb8/0x305 (Errors)

Apr  4 01:15:08 Tower kernel:  [<c1048c43>] ? unlock_page+0x18/0x1b (Errors)

Apr  4 01:15:08 Tower kernel:  [<c1057c63>] ? __do_fault+0x3a7/0x3da (Errors)

Apr  4 01:15:08 Tower kernel:  [<c105985f>] ? handle_mm_fault+0x42d/0x8f1 (Errors)

Apr  4 01:15:08 Tower kernel:  [<c108b6c6>] block_ioctl+0x2a/0x32 (Errors)

Apr  4 01:15:08 Tower kernel:  [<c108b69c>] ? block_ioctl+0x0/0x32 (Errors)

Apr  4 01:15:08 Tower kernel:  [<c10769d5>] vfs_ioctl+0x22/0x67 (Errors)

Apr  4 01:15:08 Tower kernel:  [<c1076f33>] do_vfs_ioctl+0x478/0x4ac (Errors)

Apr  4 01:15:08 Tower kernel:  [<c105dcdd>] ? do_mmap_pgoff+0x232/0x294 (Errors)

Apr  4 01:15:08 Tower kernel:  [<c1076f93>] sys_ioctl+0x2c/0x45 (Errors)

Apr  4 01:15:08 Tower kernel:  [<c1002935>] syscall_call+0x7/0xb (Errors)

Apr  4 01:15:08 Tower kernel: ---[ end trace 034abbb72e1343ab ]---

 

Sorry for the rambling guys just trying to make sure I include pertinent information... Ive attached a full syslog also though it goes back a few days.

Thanks for any help guys. 

syslog.zip

How did you use hdparm? Did you set the new size to permanent using the -p switch? Also, I'm assuming the posted syslog is after you tried to remove the HPA?

I ran hdparm -N p2930277168 /dev/sdj

I got the 2930277168 from the syslog.

The syslog is after i tried to run hdparm yes but before I rebooted the array. I shut down the server after I tried hdparm and havent restarted it since.

I can't open the syslog from my phone, but I'd restart and see if you still get the error. The message you get from running the hdparm command is misleading. When I used it I didn't think it had worked until I restarted my server.

Alright so I restarted my server and now the 1.5T disk is only recognized as ~400GB. Im guessing my hdparm number was wrong?

 

EDIT: Attached a syslog from after i restarted. As I was scrolling through it a couple things jumped out at me though honestly I have no idea what they really mean...

 

these two lines:

Apr  5 16:17:15 Tower kernel: pci 0000:00:1d.7: EHCI: BIOS handoff failed (BIOS bug?) 01010001 (Minor Issues) <--- Is this possibly a BIOS issue? Should i look into updating my BIOS?

 

Apr  5 16:17:15 Tower kernel: scst: ***WARNING***: Patch io_context was not applied on your kernel. SCST will be working with not the best performance.

syslog.txt

In your first syslog you had 3 HPA's detected, each with different reported sizes. Did you make sure you were looking at the correct one when running the hdparm command? The second syslog you posted has only two reported HPA's.

 

As far as the errors go, I actually have the second one in my syslog as well, so hopefully someone will chime in regarding that issue (shameless bump  )

 

 

The error message appears to be a result of the driver for the SASLP-MV8 not being fully optimized in this version of the kernel. I don't know if it really matters for unraid. This is total speculation based on the results I got by googling the error.

In your first syslog you had 3 HPA's detected, each with different reported sizes. Did you make sure you were looking at the correct one when running the hdparm command? The second syslog you posted has only two reported HPA's.

 

As far as the errors go, I actually have the second one in my syslog as well, so hopefully someone will chime in regarding that issue (shameless bump  )

I am pretty sure it was the right drive as it is the only 1.5Tb drive i have in the system.

The reason for the lack of the third HPA in the second syslog is because I unassigned the parity drive from the array and then restarted it in hopes that it was just that it hadnt been restarted that the hdparm didnt work. It turns out thats when it said it was only ~400GB. I tried hdat2 on the drive and it gave me an error also. then I tried the auto remove in hpa and it said it worked so I tried it in unraid again but it still says it has HPA.. I think I am just gonna sync parity and leave it for now.. I dont have any other drive in the array bigger than 1TB and I'm hoping to upgrade to a few 2TB drives in the next few months so I will just replace the parity drive then with a fresh drive.

Thanks for the help guys I appreciate it.