[Plugin] Spin Down SAS Drives


doron

Recommended Posts

25 minutes ago, Knights21 said:

No nothing and not reported on other two SAS drives on same controller.

The latter might be related to different work profiles (e.g. often-written files or folders on this sdg drive, etc.).

To eliminate that I'd swap cables / ports with the other ones and see whether issue stays with dev (i.e. sdg) or with drive (i.e. S/N / Unraid slot).

Link to comment
  • 4 weeks later...
On 9/30/2020 at 12:34 PM, doron said:

EDIT: It appears that some combinations of SAS drives / controllers are not compatible with temporary spin-down. We've seen reports specifically re Seagate Constellation ES.3 and Hitachi 10KRPM 600GB but there are probably others. Plugin has been updated to exclude combinations known to misbehave, and to use a dynamic exclusion table so that other combinations can be added from time to time. 19-Nov-2020

 

In July 21, I installed 3x 600GB SAS drives in my unraid system and installed this plugin.

 

Unfortunately, it does not work for me. I get random read/write errors on 2x drives. One of the drives just went offline while shutting down all drives while writing to the drive simultaniously.

 

I am using an LSI3008 card and 2x Toshiba AL14SEB060N 600GB drives (Unraid 6.9.2).

Interestingly, I have also 1x HGST HUC101860CSS204, which is working fine so far.

 

It seems like Unraid does not realize the drive is spun down and tries to access it regardlessly, resulting in read/write errors:

 

...
Nov  1 15:13:50 Tower emhttpd: spinning down /dev/sdp
Nov  1 15:13:50 Tower SAS Assist v0.86: Spinning down device /dev/sdp
Nov  1 15:13:59 Tower kernel: sd 15:0:7:0: [sdp] tag#4350 UNKNOWN(0x2003) Result: hostbyte=0x00 driverbyte=0x08 cmd_age=8s
Nov  1 15:13:59 Tower kernel: sd 15:0:7:0: [sdp] tag#4350 Sense Key : 0x2 [current] 
Nov  1 15:13:59 Tower kernel: sd 15:0:7:0: [sdp] tag#4350 ASC=0x4 ASCQ=0x11 
Nov  1 15:13:59 Tower kernel: sd 15:0:7:0: [sdp] tag#4350 CDB: opcode=0x2a 2a 00 28 25 b9 18 00 04 00 00
Nov  1 15:13:59 Tower kernel: blk_update_request: I/O error, dev sdp, sector 673560856 op 0x1:(WRITE) flags 0x0 phys_seg 128 prio class 0
Nov  1 15:13:59 Tower kernel: md: disk10 write error, sector=673560792
Nov  1 15:13:59 Tower kernel: md: disk10 write error, sector=673560800
...

 

...
Nov  1 15:13:59 Tower emhttpd: spinning down /dev/sdo
Nov  1 15:13:59 Tower SAS Assist v0.86: Spinning down device /dev/sdo
Nov  1 15:14:08 Tower kernel: sd 15:0:6:0: [sdo] tag#4276 UNKNOWN(0x2003) Result: hostbyte=0x00 driverbyte=0x08 cmd_age=8s
Nov  1 15:14:08 Tower kernel: sd 15:0:6:0: [sdo] tag#4276 Sense Key : 0x2 [current] 
Nov  1 15:14:08 Tower kernel: sd 15:0:6:0: [sdo] tag#4276 ASC=0x4 ASCQ=0x11 
Nov  1 15:14:08 Tower kernel: sd 15:0:6:0: [sdo] tag#4276 CDB: opcode=0x28 28 00 19 ce 1b 40 00 00 20 00
Nov  1 15:14:08 Tower kernel: blk_update_request: I/O error, dev sdo, sector 432937792 op 0x0:(READ) flags 0x0 phys_seg 4 prio class 0
Nov  1 15:14:08 Tower kernel: md: disk11 read error, sector=432937728
Nov  1 15:14:08 Tower kernel: md: disk11 read error, sector=432937736
...

 

(Edit: added Info about the HGST-Drive)

Edited by madejackson
Link to comment
  • 2 weeks later...

Hi,

Thanks for this plugin, which ran partially well on my system 6.9.1. After installing it, 12 of my 14 drives (4 sas drives ST6000NM0034) spinned down, bringing down a lot power consumption. But once spinning up, 2 of the 4 sas drives gave tens of errors, which returned to "0 disk with errors" following a reboot. Now I've removed it and never been given error warnings again. But power saving function doesn't work as efficiently as it is enchanced  by this plugin.

Link to comment

Indeed, @madejackson and @francishefeng59, unfortunately some SAS drives (mostly reported with Seagates but with some others as well) do not respond as expected to the spin-down (aka STANDBY) command and require an explicit wake-up call, which is not really applicable in the Unraid realm (expected behavior: Spin up automatically upon next i/o).

The result is typically an i/o error with sense 0x2 and ASC/ASCQ 0x4/0x11, recoverable only upon a reboot. I've made some attempts to map the bad actors (seems to be drive+controller related) and exclude them, but can't say it was an overwhelming success. So this is where we stand right now.

Link to comment
15 hours ago, doron said:

Indeed, @madejackson and @francishefeng59, unfortunately some SAS drives (mostly reported with Seagates but with some others as well) do not respond as expected to the spin-down (aka STANDBY) command and require an explicit wake-up call, which is not really applicable in the Unraid realm (expected behavior: Spin up automatically upon next i/o).

The result is typically an i/o error with sense 0x2 and SC/ASCQ 0x4/0x11, recoverable only upon a reboot. I've made some attempts to map the bad actors (seems to be drive+controller related) and exclude them, but can't say it was an overwhelming success. So this is where we stand right now.

Many thanks for the reply. Awaiting improvement. One more question. Does it affect the normal operation of unraid system or the disks if I still want it to work under error warnings?

Link to comment
Many thanks for the reply. Awaiting improvement. One more question. Does it affect the normal operation of unraid system or the disks if I still want it to work under error warnings?
If you get the i/o errors mentioned above on an array drive, the drive will need a reboot to resume operation (the message is not a warning). So, no.

Sent from my tracking device using Tapatalk

Link to comment

I have 4xHUS726060AL5210 HGST SAS drives on LSI 9200-8i HBA. On 6.9.2 they are can't be spin down. On 6.9.1 it works perfect but for a couple of reasons I can't downgrade to 6.9.1. It there a fix for this? Is this Unraid issue or plugin issue?

 

UPD: Same issue on 6.10-rc2.

Edited by SuberSeb
Link to comment

New to UnRAID...  man this is awesome.     Days later --- hhmmm disk are not spinning down.    Cache folder setup maybe...    hhhmmm   still no spin down.

Find this APP --- install --   disk spin down as expected.

 

Awesome!  Just wanted to say thank for the plugin 😃

 

**All WD RE and Gold drives here so no issues

Edited by Brion
Link to comment
On 11/16/2021 at 10:05 AM, SuberSeb said:

I have 4xHUS726060AL5210 HGST SAS drives on LSI 9200-8i HBA. On 6.9.2 they are can't be spin down. On 6.9.1 it works perfect but for a couple of reasons I can't downgrade to 6.9.1. It there a fix for this? Is this Unraid issue or plugin issue?

 

UPD: Same issue on 6.10-rc2.

Do you see a "SAS Assist" message about spinning the drive down, and then immediately thereafter a message about reading SMART?

 

There was an issue that started in 6.9.2 (vs. 6.9.1), where some drives are spinning down and immediately back up. It seems to be a kernel issue, I'm not sure it's been resolved yet. It was reported against both SAS and SATA drives.

Link to comment
17 hours ago, doron said:

Do you see a "SAS Assist" message about spinning the drive down, and then immediately thereafter a message about reading SMART?

Exactly.

Nov 19 09:08:39 R7UNRAID emhttpd: spinning down /dev/sdi
Nov 19 09:08:39 R7UNRAID SAS Assist v0.86: Spinning down device /dev/sdi
Nov 19 09:08:45 R7UNRAID emhttpd: read SMART /dev/sdi

 

17 hours ago, doron said:

There was an issue that started in 6.9.2 (vs. 6.9.1), where some drives are spinning down and immediately back up. It seems to be a kernel issue, I'm not sure it's been resolved yet. It was reported against both SAS and SATA drives.

Seems that is my case... Also this happening on 6.10-rc2.

Link to comment
  • 1 month later...
I am having the same issue. It's a known issue. Fixes are probably coming, either with the plugin or unraid kernel.
Just to clarify, this appears to be unrelated to the plugin, and applies to both SATA and SAS drives. Those drives (mainly Seagates) are spun up soon after being spun down. Seems to be kernel related, though not sure.

Sent from my tracking device using Tapatalk

Link to comment
36 minutes ago, doron said:

Just to clarify, this appears to be unrelated to the plugin, and applies to both SATA and SAS drives. Those drives (mainly Seagates) are spun up soon after being spun down. Seems to be kernel related, though not sure.

Sent from my tracking device using Tapatalk
 

Ok...most of my drives are HGST and it still happens, so it appears to be kernel related then.

Link to comment
Ok...most of my drives are HGST and it still happens, so it appears to be kernel related then.
Quite an elusive issue. My drives are old WD and newer HGST and I don't see that problem at all, neither in 6.9 nor in 6.10-rc1/2.

Sent from my tracking device using Tapatalk

Link to comment
  • 3 weeks later...

Posting here for additional info as it currently seems to be the most relevant place.

 

Getting the issue with 6.9.2 and 6.10.0-rc2.

I've tried the FCP plugin, and I'm using the latest SAS spin down plugin. Before the plugin my drives wouldn't spin down, now they do, only to be met very shortly after with a 'emhttpd: read SMART /dev/...' message.

 

I've got no other services running and there should be no reason any drives are accessed, data is only stored on the first drive.

 

I have 2 drives in parity, another 19 in the main array. All 1.2TB SAS HPE drives, with a mix of three different models. Same issue across any model.

 

Drives are working correctly. Setup is a HP D2700 (25 2.5" drives, 21 1.2TB, 4 500GB) to a HP P222 in HBA mode. P222 is in a HP DL380P G8.

 

I've gone through a lot of hurdles, this is my last. Once I can power down most of my drives (the D2700 is new) I will save power and noise for quite a while.

 

If anyone needs diagnostics let me know.

Edited by peachyojon
Link to comment
  • 4 weeks later...
  • 2 weeks later...

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.

Guest
Reply to this topic...

×   Pasted as rich text.   Restore formatting

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.