storage-drives

This content is provided under the GNU Free Documentation License, Version 1.3, 3 November 2008

NONE OF THIS IS COMPUTER ADVICE.

Please do not set fire to your processor, break the Computer Misuse Act, or harm others’ computers.

THESE CHICKEN SCRATCHES ARE PROVIDED “AS IS”, WITHOUT WARRANTY OF ANY KIND, EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE CONTENT HEREIN.

Tools:

SATA and – probably even moreso – SCSI are things you ought to RTFM for.

LSI/SAS/HBA/RAID cards

I experienced a world of hurt attempting to get a SAS3008 card with its default firmware (supporting RAID+JBOD modes) to work.

[    4.297045] megaraid_sas 0000:01:00.0: FW now in Ready state
[    4.297048] megaraid_sas 0000:01:00.0: 63 bit DMA mask and 32 bit consistent mask
[    4.297056] mc: Linux media interface: v0.10
[    4.297304] megaraid_sas 0000:01:00.0: firmware supports msix        : (96)
[    4.302519] iTCO_vendor_support: vendor-support=0
[    4.307340] PTP clock support registered
[    4.311236] megaraid_sas 0000:01:00.0: requested/available msix 9/9 poll_queue 0
[    4.311244] megaraid_sas 0000:01:00.0: current msix/max num queues   : (9/8)
[    4.311247] megaraid_sas 0000:01:00.0: RDPQ mode     : (disabled)
[    4.311250] megaraid_sas 0000:01:00.0: Current firmware supports maximum commands: 272        LDIO threshold: 237
[    4.312327] megaraid_sas 0000:01:00.0: Performance mode :Latency (latency index = 1)
[    4.312332] megaraid_sas 0000:01:00.0: FW supports sync cache        : No
[    4.312336] megaraid_sas 0000:01:00.0: megasas_disable_intr_fusion is called outbound_intr_mask:0x40000009
[    4.320322] iTCO_wdt iTCO_wdt.1.auto: Found a Lynx Point TCO device (Version=2, TCOBASE=0x1860)
[    4.322563] iTCO_wdt iTCO_wdt.1.auto: initialized. heartbeat=30 sec (nowayout=0)
[    4.326624] e1000e: Intel(R) PRO/1000 Network Driver
[    4.326627] e1000e: Copyright(c) 1999 - 2015 Intel Corporation.
[    4.332843] e1000e 0000:00:19.0: Interrupt Throttling Rate (ints/sec) set to dynamic conservative mode
[    4.333399] megaraid_sas 0000:01:00.0: Init cmd return status FAILED for SCSI host 6
[    4.343977] megaraid_sas 0000:01:00.0: Failed from megasas_init_fw 6553

My eventual “fix” was to give up and flash IT mode firmware, but some of the diagnostic steps I tried that may work for you:

Booting with kernel parameter combinations:
- iommu=on
- iommu=pt
- iommu=soft
- iommu=on intel_iommu=pt
- iommu=pt intel_iommu=on
Booting in CSM mode
Booting in UEFI mode
Factory resetting BIOS
Factory resetting SAS card
Changing SAS card’s config of exposed drives to host
Setting all drives as “JBOD” in SAS card
Setting all drives as “unconfigured” for RAID mode in SAS card
Creating RAID array in SAS card
Unplugging drives and/or SFF8643/SFF8087 cable
Changing PCIe slot
Booting Ubuntu 24.04
Booting Windows 10 (got error Code 10 in Device Manager)
Booting Ubuntu in a different machine (got megasas_init_fw 6550 instead, but still not working)

Drive and File activity

See files being accessed:

lsof | grep /folder

See disk activity

iostat -x 1

SATA spin down

This sets drive to spin down after 5m

https://bbs.archlinux.org/viewtopic.php?id=216832

sudo hdparm -B 128 -S 60 /dev/sdd

SAS EPC settings

References:

Ensure you know what the idle/standby states do and that you know what you are doing.

This is used instead of APM on SAS drives.

The enabled modes will have an asterisk by them. This is important. If none are enabled, the drive won’t spin down.

For example, here standby_z is the only enabled state:

==========================================================================================
 openSeaChest_PowerControl - openSeaChest drive utilities - NVMe Enabled
 Copyright (c) 2014-2025 Seagate Technology LLC and/or its Affiliates, All Rights Reserved
 openSeaChest_PowerControl Version: 3.7.2 X86_64
 Build Date: Mar  6 2026
 Today: 20260307T120113 User: root
==========================================================================================

/dev/sde - ST6000NM0034   X - S4D0TSBZ0000K64564C4 - BC6G - SCSI

===EPC Settings===
        * = timer is enabled
        C column = Changeable
        S column = Savable
        All times are in 100 milliseconds

Name       Current Timer Default Timer Saved Timer   Recovery Time C S
Idle A      1200          10            1200         1             Y Y
Idle B      2400          2400          2400         3             Y Y
Idle C      6000          6000          6000         70            Y Y
Standby Y   6000          6000          6000         70            Y Y
Standby Z  *9000          9000          9000         150           Y Y

Power Condition Name	Description
Idle_a	Reduced electronics
Idle_b	Heads unloaded. Disks spinning at full RPM
Idle_c	Heads unloaded. Disks spinning at reduced RPM
Standby_z	Heads unloaded. Motor stopped (disks not spinning)

Explanation of the idle/standby states grabbed from https://www.seagate.com/www-content/product-content/enterprise-hdd-fam/exos-x-16/en-us/docs/100845789j.pdf

The plastic of my keycaps has rubbed off on my fingers and I still haven’t figured out a reliable way to get different types of drives to just spin down… I wish better luck to your ventures than I have had.

Show info and settings

sudo openSeaChest_Info -d /dev/sde -i
sudo openSeaChest_PowerControl -d /dev/sde --showEPCSettings
sudo sdparm --flexible -6 -l --all /dev/sde

Configure spindown

The short and sweet explanation:

EPC may be enabled/disabled as a whole. It needs to be enabled
Each individual idle/standby state may be enabled/disabled independently. I do not know the implications, if any, of enabling one but not another.
Each idle/standby state will have a different timer. Be wary of setting the times too short, know your environment.
The time units are in 100ms, e.g., “9000” == 900000ms == 15m
What works for one drive may not work for another, and perhaps nothing may work for some drives (the bitter part)

I was getting errors when using the openSeaChest_PowerControl functions

The sdparm functions worked for me

If you need to, enable EPC:

sudo openSeaChest_PowerControl -d /dev/sde --EPCfeature enable

Change the time to enter a given idle/standby state (openSeaChest, didn’t work for me):

sudo openSeaChest_PowerControl -d /dev/sde --standby_z 9000

Enable an idle/standby state:

This worked for my “IBM-ESXS” ST6000NM0034 X drive, but not my Seagate DKS2F-H6R0SS drive

sudo sdparm --flexible -6 -l --set standby_z=1 /dev/sde

If these options still do not work, and you have two identical drives, with one misbehaving, you could try and compare their settings (sdparm --all /dev/sde). Albeit to no avail, I changed the PERF and PM attributes in one such attempt.

Force SAS spindown manually

The openSeaChest command spun the drive down but it immediately spun back up

sudo openSeaChest_PowerControl -d /dev/sde --spinDown

This worked but the drive doesn’t come back up automatically

sudo sdparm --command=stop /dev/sde