Disk Spin up

On 3/14/2022 at 10:39 PM, dlandon said:

Thank you.  Unraid is in fact adding the parity disk to the unassigned disks,  I'll report this and see if we can get a resolution.

Any feedback so far?

Nothing yet.

We haven't been able to reproduce the issue.  I noticed you have disk slots that are not assigned to disks.  Can you re-do the disk slots to just what is needed for your array?  This includes array and pool devices.  And give it another shot.

6 hours ago, dlandon said:

We haven't been able to reproduce the issue.  I noticed you have disk slots that are not assigned to disks.  Can you re-do the disk slots to just what is needed for your array?  This includes array and pool devices.  And give it another shot.

Change slot count to only include active drives, removed secondpool which didn't have a disk in it.

 

Spun down disks added flash drive and ran /usr/local/sbin/emcmd 'cmdHotplug=apply' as UD is not installed.

 

Same results.

 

Just dev 1 and 2 existed before now 4 show dev4 is parity. Fresh diags attached.

 

 

Also there is another post of similar issue.

 

 

tower-diagnostics-20220325-0633.zip

Would you be receptive to a Team Viewer session or access to your server so I can get a first hand look?

1 hour ago, dlandon said:

Team Viewer session

Yes would be happy to do that when would be a good date and time?

Next week some time.  I need to organize what I need to test and what information I need to collect so we can get this solved.

 

I'll get back to you on a date and time.  Where are you located so I can take into account the time difference?

UK

I'm in the US Central Time Zone.  So 5 hours ahead of me?

Yes but that will change on Sunday as we go into British Summer Time Currently 1226 here next week will be 1326

I will probably want to do early morning here.  7 or 8 AM.

 

I've asked Tom what tests/data he would like me to do/collect so he can replicate the issue.  I don't have SAS hardware so I can't reproduce the problem.  He has SAS hardware so he should be able to reproduce.

Ok that is going to be in the middle of my work day, but Fridays I can do afternoon my time as finish at 12.15 as long as no issues.

I could do later so it's your evening if that works better.  I do not have a structured job, so I'm very flexible.

Would Sunday this week work better for you?

Yes I could do Sunday.

Ok i think the version of smartctl currently being used doesnt have the fix of 2021-01-24 I have rebooted and parity is now sdg.

 

This was added as of r5179 and current smartctl version is r5155 below is testing with r5215 i created for testing in April 2021

 

So looks like the hotplug is not calling smartctl as the issue still appears with the new version.

 

Mar 27 15:36:49 Tower unassigned.devices: Processing Hotplug event...
Mar 27 15:36:56 Tower kernel: sd 3:0:1:0: attempting task abort!scmd(0x00000000d7956a2d), outstanding for 7060 ms & timeout 7000 ms
Mar 27 15:36:56 Tower kernel: sd 3:0:1:0: [sdg] tag#7477 CDB: opcode=0x12 12 01 00 00 fe 00
Mar 27 15:36:56 Tower kernel: scsi target3:0:1: handle(0x000e), sas_address(0x5000c500626c01e9), phy(2)
Mar 27 15:36:56 Tower kernel: scsi target3:0:1: enclosure logical id(0x5003048011c7eb00), slot(1) 
Mar 27 15:36:56 Tower emhttpd: ST4000NM0023_35000c500626c01eb (sdg) 512 7814037168
Mar 27 15:36:56 Tower kernel: sd 3:0:1:0: task abort: SUCCESS scmd(0x00000000d7956a2d)
Mar 27 15:36:56 Tower emhttpd: read SMART /dev/sdg

 

https://www.smartmontools.org/changeset/5179

 

Below shows issue on r5155 and  no issue on the r5215 vers, Note that this fix, fixes issue for Seagate SAS drives.

 

root@Tower:/usr/local/sbin# lsscsi
[0:0:0:0]    disk    SanDisk  Cruzer Fit       1.00  /dev/sda 
[1:0:0:0]    disk    SanDisk  Cruzer Blade     1.00  /dev/sdb 
[2:0:0:0]    disk    USB 2.0  Flash Disk       2.00  /dev/sdc 
[3:0:0:0]    disk    HITACHI  HMRSK2000GBAS07K 3P02  /dev/sdf 
[3:0:1:0]    disk    SEAGATE  ST4000NM0023     XMGJ  /dev/sdg 
[3:0:2:0]    disk    HGST     HUS724030ALS640  A1C4  /dev/sdh 
[3:0:3:0]    disk    HGST     HUS724030ALS640  A1C4  /dev/sdi 
[3:0:4:0]    disk    HGST     HUS724030ALS640  A1C4  /dev/sdj 
[3:0:5:0]    disk    HGST     HUS724030ALS640  A1C4  /dev/sdk 
[6:0:0:0]    disk    ATA      SSD PLUS 480GB   00RL  /dev/sdd 
[8:0:0:0]    disk    ATA      CT500MX500SSD1   023   /dev/sde 
[10:0:0:0]   disk    HITACHI  HMRSK2000GBAS07K 3P02  /dev/sdl 
[10:0:1:0]   disk    ATA      MB0500GCEHE      HPG1  /dev/sdm 
root@Tower:/usr/local/sbin#  smartctl -n standby /dev/sdg
smartctl 7.2 2020-12-30 r5155 [x86_64-linux-5.15.30-Unraid] (local build)
Copyright (C) 2002-20, Bruce Allen, Christian Franke, www.smartmontools.org

Device is in ACTIVE mode
SCSI device successfully opened

Use 'smartctl -a' (or '-x') to print SMART (and more) information

root@Tower:/usr/local/sbin#  smartctl.r5215 -n standby /dev/sdg
smartctl 7.3 2021-04-07 r5215 [x86_64-linux-5.15.30-Unraid] (CircleCI)
Copyright (C) 2002-21, Bruce Allen, Christian Franke, www.smartmontools.org

Device is in ACTIVE mode
SCSI device successfully opened

Use 'smartctl -a' (or '-x') to print SMART (and more) information

root@Tower:/usr/local/sbin#  smartctl.r5215 -s standby,now /dev/sdg
smartctl 7.3 2021-04-07 r5215 [x86_64-linux-5.15.30-Unraid] (CircleCI)
Copyright (C) 2002-21, Bruce Allen, Christian Franke, www.smartmontools.org

Device placed in STANDBY mode
SCSI device successfully opened

Use 'smartctl -a' (or '-x') to print SMART (and more) information

root@Tower:/usr/local/sbin#  smartctl.r5215 -n standby /dev/sdg
smartctl 7.3 2021-04-07 r5215 [x86_64-linux-5.15.30-Unraid] (CircleCI)
Copyright (C) 2002-21, Bruce Allen, Christian Franke, www.smartmontools.org

Device is in STANDBY BY COMMAND mode, exit(2)
root@Tower:/usr/local/sbin# 

 

smartctl change log.

 

2021-01-24  Simon Fairweather  

	scsiprint.cpp: Move VPD page process to after new powermode processing
	added in r5131 to resolve #1413 of Seagate drives spinning up when
	-n standby is selected.  Simplify code for VPD pages.

2020-12-15  Douglas Gilbert  <dgilbert@interlog.com>

	smartctl: expand -s option with standby,now and standby,off (or
	standby,0) to include SCSI. Modified code from Simon Fairweather
	found in github pull #72. As per my 20201205 patch, this
	area (i.e. SCSI power conditions including START and STOP) needs
	to be revisited; leave that until after the 7.2 release.


2020-12-05  Douglas Gilbert  <dgilbert@interlog.com>

	smartctl: expand -n option to include SCSI. Code from Simon
	Fairweather. Still thinking about how to handle SCSI "stopped"
	state which requires the user to send a SCSI command to restart.

 

Edited March 27, 2022 by SimonF

Smartctl updated to 7.3 in next release and this issue should be fixed in 6.10rc5, tested on internal release, will wait for RC5 to drop before closing this issue.

Changed Status to Retest

Changed Status to Solved

2 hours ago, SimonF said:

Changed Status to Solved

Hallelujah!