| View previous topic :: View next topic |
| Author |
Message |
omega13333
n00b
Joined: 23 Oct 2010
Posts: 5
|
 Posted: Mon Nov 22, 2010 6:41 pm Post subject: |
 |
|
My unit is accessible about 10 seconds after wakeup from standby. Im running 3 WD-1TB-drives in a RAID5 setup and my operating system (debian) runs inside a LVM volume on the raid itself. (I'm not using a dedicated drive at the IDE-port at the moment, because I have not found time to migrate it) Therefor my drives also don't spin down, but I've configured my NAS to go back to standby after 30 min of inactivity.
I can't tell you anything about speedstep because I have not changed the celeron CPU. |
|
| Back to top |
|
 |
buzzone
n00b
Joined: 31 Jan 2011
Posts: 3
|
| Posted: Mon Jan 31, 2011 1:04 pm Post subject: |
|
|
I have got the SS4200 hardware and am using 4 SATA drives and one PATA drives with Gentoo installed.
I had exactly the same problem that the PATA drive was not recognized by the kernel. Then I found this thread applied the patch, recompiled. Now the PATA drive is recognized:
| Code: |
[ 4.559614] ata5: PATA max UDMA/100 cmd 0x1f0 ctl 0x3f6 bmdma 0xffa0 irq 14
[ 4.724088] ata5.00: ATA-8: SAMSUNG HM160HC, LQ100-10, max UDMA/100
[ 4.724093] ata5.00: 312581808 sectors, multi 16: LBA48
[ 4.732066] ata5.00: configured for UDMA/100
|
This is a 2.5 inch notebook PATA HDD. I am using a brand-new modern 80-pin IDE cable and then a 3.5" to 2.5" adapter. What happens now about once every five minutes or so is this:
| Code: |
[ 4567.000062] ata5: lost interrupt (Status 0x58)
[ 4567.000112] ata5.00: exception Emask 0x0 SAct 0x0 SErr 0x0 action 0x6 frozen
[ 4567.000117] ata5.00: BMDMA stat 0x4
[ 4567.000122] ata5.00: failed command: WRITE DMA
[ 4567.000132] ata5.00: cmd ca/00:08:6d:f3:62/00:00:00:00:00/e3 tag 0 dma 4096 out
[ 4567.000134] res 58/00:08:6d:f3:62/00:00:00:00:00/e3 Emask 0x2 (HSM violation)
[ 4567.000140] ata5.00: status: { DRDY DRQ }
[ 4567.000173] ata5: soft resetting link
[ 4567.164069] ata5.00: configured for UDMA/100
[ 4567.164081] ata5: EH complete
|
I tried to manually set a lower UDMA mode using hdparm. But I can't:
| Code: |
$ sudo hdparm -d -X udma3 /dev/sde
/dev/sde:
setting xfermode to 67 (UltraDMA mode3)
HDIO_DRIVE_CMD(setxfermode) failed: Input/output error
HDIO_GET_DMA failed: Inappropriate ioctl for device
|
I tried to set the UDMA mode in the BIOS to a lower value (UDMA3). But Linux still sets it to UDMA5 always.
It can't really check the HDD because I don't have replacement hardware right now. Maybe it is the setup with my cable/adapter or the HDD is really failing?
Does somebody else experience such behavior using the patch? Probably it doesn't have anything to do with the patch? |
|
| Back to top |
|
|
roarinelk
Guru
Joined: 04 Mar 2004
Posts: 520
|
| Posted: Mon Jan 31, 2011 9:34 pm Post subject: |
|
|
either get shorter cables or don't use the ones with 80 wires.
shorter cables mean less surface for EMI; using 40 wire cables
should make the ata driver omit the >=udma5 modes when it does
cable detection.
hdparm is useless with a modern 2.6 kernel btw. |
|
| Back to top |
|
|
VoidMage
Watchman
Joined: 14 Oct 2006
Posts: 6196
|
| Posted: Mon Jan 31, 2011 10:03 pm Post subject: |
|
|
| roarinelk wrote: | | hdparm is useless with a modern 2.6 kernel btw. |
And why is that ? |
|
| Back to top |
|
|
roarinelk
Guru
Joined: 04 Mar 2004
Posts: 520
|
| Posted: Mon Jan 31, 2011 10:11 pm Post subject: |
|
|
| I meant to say "not useful for setting DMA modes". it still has some other uses of course (like spindown time, ...) |
|
| Back to top |
|
|
buzzone
n00b
Joined: 31 Jan 2011
Posts: 3
|
| Posted: Sat Feb 05, 2011 11:00 pm Post subject: |
|
|
great! 
and thanks! i'll go for 40-pin/shorter cables. will report on the results then. |
|
| Back to top |
|
|
buzzone
n00b
Joined: 31 Jan 2011
Posts: 3
|
| Posted: Mon Feb 28, 2011 11:39 am Post subject: |
|
|
just for the archive:
i tried a 40-pin cable. the hdd was detected by the bios but didn't show up as a boot device. (was a very old cable. presumably broken)
but i was switching now to another solution anyway: i used a compact flash ide adapter that sits directly on the mainboard. now i have a read-only root filesystem on the CF with a rw tmpfs on top (using aufs2) which works just great.
i have no issues anymore with "soft resetting link". also the hdd works perfectly in another setup. so i'm quite confident that it was EMI because of the long IDE cable.
thanks everyone for the replies.
i consider this post solved/closed for now. |
|
| Back to top |
|
|
garyn_87048
n00b
Joined: 28 Oct 2013
Posts: 1
|
| Posted: Mon Oct 28, 2013 12:11 am Post subject: |
|
|
Are there any updates to this project? I realize this thread is about 3yr old. I also have an ss4200 that I would like to run FreeNAS on. I'm relatively new to linux, but I would like to build a version of FreeNAS to work in this box. I recently purchased a 4G DOM to use for this project.
The thread mentioned a possible BIOS update for the ss4200. Any chance that occurred (wishful thinking!!!)?
Thanks!
Gary |
|
| Back to top |
|
|
|
Kernel & Hardware
