On 1/10/2025 at 4:47 PM, Jestern said:

I can confirm the package from Slackware current, as of today, seems to work fine on Unraid 7.0.0
 

https://mirrors.slackware.com/slackware/slackware64-current/slackware64/ap/powertop-2.15-x86_64-2.txz

 

I can get down to C8 like before.

 

As a side note, I also tested running powertop on a privileged Ubuntu container and it seems to work fine as well, adjusting everything on the host system

I can only get to C8 on this version of powertop and unraid 7. On previous version I was able to get to C10.

 

This is the output when running powertop or any of it's associated commands

 

modprobe cpufreq_stats failed
Loaded 0 prior measurements
Cannot load from file /var/cache/powertop/saved_parameters.powertop
File will be loaded after taking minimum number of measurement(s) with battery only
RAPL device for cpu 0
RAPL device for cpu 0
Devfreq not enabled
glob returned GLOB_ABORTED
Cannot load from file /var/cache/powertop/saved_parameters.powertop
File will be loaded after taking minimum number of measurement(s) with battery only
Leaving PowerTOP

Edited January 11, 20251 yr by xreyuk

30 minutes ago, xreyuk said:

This is the output when running powertop or any of it's associated commands

Not relevant 

I recently switched to a new MB and now it seems I can't reach the same high C-States anymore.

New Mainboard is the Z790 Aorus Elite DDR4, other than that I only added an LSI-9207-8i to my system.

 

Settings are these: Link to image with settings (from this thread), since someone else with the DDR5 version of the board was able to reach C10 I was hoping I could as well with that board

 

Not sure what exactly to make of the info powertop shows me, my best guess would be something between my CPU and the LSI card I have? (Though other things have Usage at 100% too)

 

Tunables are all on good

 

The command showing ASPM states shows some as disabled, I've double-checked the bios settings, but there doesn't seem to be anything related to ASPM on disabled/auto anymore

 

Hey guy´s,

 

i have a small problem with powertop and i thought my mainboard is the problem, because with a gigabyte Mainboard and TruenasScale it works well.

But i´ve already change my Z690 Taichi to a Gigabyte b760 aorus elite ax and the Problem is still the same:

 

I only get "Pckg OS instead of Pck HW" - is there a BIOS Setting i´m missing?

 

Other Screenshot:

 

Devices:

 

The Mellanox card will be changed to a X710-DA2 - but what is the device 00:1c:4?

 

Mainboard: Gigabyte b760 aorus elite ax

 

Best regards 

Marc

 

11 minutes ago, dezai said:

is there a BIOS Setting i´m missing?

No, the used powertop version does not recognize your setup.

 

2 hours ago, OddMagnet said:

only added an LSI-9207-8i to my system.

 Test it without this card and you probably see better c-states.

Hello, following this thread, to try to configure my new UnRaid in the best way.

It is about:

 

Intel 14500
Asrock Z790 ITX/TB4
64gb ddr5 ram
2 x WD SN850X nvme for cache
3 x Sata hdd for array

M2 x6 Sata ASM1166

 

I have enabled all C states in the bios, I have also disabled thunderbolt4 (although it continues to appear in the list of devices despite being disabled) I have also disabled the sound card, wifi and some other things that are not used.

After starting UnRaid, I see C3 states on the packet, but after about an hour, UnRaid stops working (no local access to it or any of the virtual machines or dockers)

I will update with photos of the bios later so you can see my configuration and what I did wrong.

 

All the devices when listing them, I can see that they support ASPM, except these two:

 

 PCI bridge: Intel Corporation Thunderbolt 4 Bridge [Maple Ridge 4C 2020] (rev 02) (prog-if 00 [Normal decode])
                LnkCap: Port #0, Speed 8GT/s, Width x4, ASPM L0s L1, Exit Latency L0s <2us, L1 <4us
                LnkCtl: ASPM Disabled; LnkDisable- CommClk+

 

PCI bridge: Intel Corporation Raptor Lake PCI Express Root Port #21 (rev 11) (prog-if 00 [Normal decode])
                LnkCap: Port #21, Speed 16GT/s, Width x4, ASPM L1, Exit Latency L1 <64us
                LnkCtl: ASPM Disabled; RCB 64 bytes, LnkDisable- CommClk+

 

 

 

Thank you above all for your time and help.

Edited January 14, 20251 yr by Nozle
Upload images

On 1/14/2025 at 8:45 AM, Nozle said:

Hello, following this thread, to try to configure my new UnRaid in the best way.

It is about:

 

Intel 14500
Asrock Z790 ITX/TB4
64gb ddr5 ram
2 x WD SN850X nvme for cache
3 x Sata hdd for array

M2 x6 Sata ASM1166

 

I have enabled all C states in the bios, I have also disabled thunderbolt4 (although it continues to appear in the list of devices despite being disabled) I have also disabled the sound card, wifi and some other things that are not used.

After starting UnRaid, I see C3 states on the packet, but after about an hour, UnRaid stops working (no local access to it or any of the virtual machines or dockers)

I will update with photos of the bios later so you can see my configuration and what I did wrong.

 

All the devices when listing them, I can see that they support ASPM, except these two:

 

 PCI bridge: Intel Corporation Thunderbolt 4 Bridge [Maple Ridge 4C 2020] (rev 02) (prog-if 00 [Normal decode])
                LnkCap: Port #0, Speed 8GT/s, Width x4, ASPM L0s L1, Exit Latency L0s <2us, L1 <4us
                LnkCtl: ASPM Disabled; LnkDisable- CommClk+

 

PCI bridge: Intel Corporation Raptor Lake PCI Express Root Port #21 (rev 11) (prog-if 00 [Normal decode])
                LnkCap: Port #21, Speed 16GT/s, Width x4, ASPM L1, Exit Latency L1 <64us
                LnkCtl: ASPM Disabled; RCB 64 bytes, LnkDisable- CommClk+

 

 

 

Thank you above all for your time and help.

 

Did you tune with Powertop? Your Intel Networkcard won't support any energy savings and if you tune it with Powertop the System gets unresponsive. 

https://bugzilla.kernel.org/show_bug.cgi?id=218499

Edited January 17, 20251 yr by FlowDE

On 1/17/2025 at 12:22 PM, FlowDE said:

 

Did you tune with Powertop? Your Intel Networkcard won't support any energy savings and if you tune it with Powertop the System gets unresponsive. 

https://bugzilla.kernel.org/show_bug.cgi?id=218499

Hello, thank you very much for your reply and information, I didn't know that the version of this ethernet is 226V, it only hangs if I activate C7 in the bios (if I enable all ASPM and C states except C7) everything is fine... But I only get to C3 in package and c6 in core (c2_acpi in cpu) maximum.

If I enable C7 in bios, I get to C3 in package, c7 in Core, C3_ACPI maximum.

So, I'm not very clear now, what is the solution to avoid it hanging with this 226V?

 

I didn't do anything additional in UnRaid (it doesn't matter if I run the command: powertop --auto-tune &>/dev/null or I don't run it) it hangs whenever c7 is enabled in the bios, it doesn't matter if I run that command or not.

On the other hand: Should PCI EXPRESS NATIVE CONTROL be enabled or disabled?

In the motherboard manual it says that the network card is: E3100X 2.5G LAN

 

lspci:

 

09:00.0 Ethernet controller [0200]: Intel Corporation Device [8086:3102] (rev 04)
 

Edited January 18, 20251 yr by Nozle

2 hours ago, Nozle said:

E3100X 2.5G LAN

As this controller was released in 2021 it shouldn't be the I226V, but finally it could be still an issue with this controller or anything else on the board or... Did you update the firmware of the ASM1166? Older firmware versions had problems with deep sleeps states, but usually they didn't cause a system crash, "only" disconnected disks.

27 minutes ago, mgutt said:

As this controller was released in 2021 it shouldn't be the I226V, but finally it could be still an issue with this controller or anything else on the board or... Did you update the firmware of the ASM1166? Older firmware versions had problems with deep sleeps states, but usually they didn't cause a system crash, "only" disconnected disks.

Yeah; The firmware was updated. Right now the bios is configured like this:

 

 

I haven't run autotune powertop, but the system hangs after about an hour. This is generating many parity errors for me.

 

unraid 7 is installed.

 

asm1166 no have disk connect (is update firmware)

Edited January 18, 20251 yr by Nozle

I tried without starting the array, and it worked normally for at least two hours (I restarted it to test with the array started), then it crashed again as usual (bios configuration as in the screenshots above)

The array is two NVME cache disks (one plugged into the chipset slot and another in the rear slot) and 3 8TB HDD disks

 

I run out of ideas if everything is configured correctly and if I'm doing it wrong or how to solve all this.

My parity every time the system hangs needs to be checked (12 hours) and found more than 1000 errors...

Did you check your logs? Maybe it is something completely different. Or an Unraid 7 bug?! Did you try a downgrade?

11 minutes ago, mgutt said:

Did you check your logs? Maybe it is something completely different. Or an Unraid 7 bug?! Did you try a downgrade?

I seem to remember that I had the same problem on unraid 6, but I couldn't be sure. I tried to check the logs but they are lost on reboot, I enabled the syslog server but there was nothing there

2 hours ago, mgutt said:

Did you check your logs? Maybe it is something completely different. Or an Unraid 7 bug?! Did you try a downgrade?

The log doesn't really say anything, the last thing I can read before it crashes is:

 

 

Jan 18 21:57:30 Tower kernel: Bluetooth: Core ver 2.22
Jan 18 21:57:30 Tower kernel: NET: Registered PF_BLUETOOTH protocol family
Jan 18 21:57:30 Tower kernel: Bluetooth: HCI device and connection manager initialized
Jan 18 21:57:30 Tower kernel: Bluetooth: HCI socket layer initialized
Jan 18 21:57:30 Tower kernel: Bluetooth: L2CAP socket layer initialized
Jan 18 21:57:30 Tower kernel: Bluetooth: SCO socket layer initialized
Jan 18 22:10:33 Tower emhttpd: spinning down /dev/sdd
Jan 18 22:10:33 Tower emhttpd: spinning down /dev/sdb
Jan 18 22:10:33 Tower emhttpd: spinning down /dev/sdc

The bios configured as in the previous photos, and powertop autotune is not executed, the array is executed.

Edited January 18, 20251 yr by Nozle

Kernel errors can happen much further. Two hints:

 

1.) Install this:

https://forums.unraid.net/topic/121039-syslog-notify-create-notifications-if-specific-words-occur-in-the-logs/

 

2.) If it becomes unresponsive and does not restart: Connect a monitor and disable monitor sleep. When a crash happens, the monitor writes the kernel error to the display. 

2 hours ago, mgutt said:

Kernel errors can happen much further. Two hints:

 

1.) Install this:

https://forums.unraid.net/topic/121039-syslog-notify-create-notifications-if-specific-words-occur-in-the-logs/

 

2.) If it becomes unresponsive and does not restart: Connect a monitor and disable monitor sleep. When a crash happens, the monitor writes the kernel error to the display. 

I ran the script from the run script plugin, and I enabled the screen, it hung again but nothing appeared on the screen (only login) and the script did not send me any email.

 

 

 

Edited January 19, 20251 yr by Nozle

1 hour ago, Nozle said:

it hung again but nothing appeared on the screen

This is strange. Are you able to login through local display / keyboard if it hangs (so "only" Network is dead) or is it really frozen?

 

Are you able to test the CPU on a different Mainboard?

6 hours ago, mgutt said:

This is strange. Are you able to login through local display / keyboard if it hangs (so "only" Network is dead) or is it really frozen?

 

Are you able to test the CPU on a different Mainboard?

So I did a clean reinstallation of unraid 7, without configuring array, or anything else, and I got a first crash, nothing in the log, nothing on the unraid screen (powertop didn't even want to be installed) I'm going crazy with this
 

no is posible login in screen Unraid, is frozen.

 

Cpu is okay in other mb.

 

If I don't enable c7 in the bios, everything is fine, the crashes appear when enabling c7 as in the photos above

 

 

PSU is Seasonic Focus SPX-650

Edited January 19, 20251 yr by Nozle

On 1/19/2025 at 11:07 AM, mgutt said:

This is strange. Are you able to login through local display / keyboard if it hangs (so "only" Network is dead) or is it really frozen?

 

Are you able to test the CPU on a different Mainboard?

I have tried a new 14500 and same problem here, it keeps hanging when enabling C7 in the bios (after about an hour)

50 minutes ago, Nozle said:

On 1/19/2025 at 11:07 AM, mgutt said:

 

I have tried a new 14500 and same problem here, it keeps hanging when enabling C7 in the bios (after about an hour)

Can you please try it with Ubuntu? Booting without installing from USB should be sufficient.

9 minutes ago, mgutt said:

Can you please try it with Ubuntu? Booting without installing from USB should be sufficient.

Sure, I'll try it! What configuration should I do? install powertop and run powertop autotune?

I was able to disable Intel LAN in the bios and get the same crash after an hour and a half.

1 hour ago, Nozle said:

Sure, I'll try it! What configuration should I do? install powertop and run powertop autotune?

As you said it crashes without powertop autotune, this shouldn't be necessary, but after 1-2 hours not crashing I would try it with powertop, too.

10 hours ago, mgutt said:

As you said it crashes without powertop autotune, this shouldn't be necessary, but after 1-2 hours not crashing I would try it with powertop, too.

After approximately two hours with ubuntu live usb and powertop autotune running, it has crashed, I am now testing without autotune.

 

14 hours ago, mgutt said:

As you said it crashes without powertop autotune, this shouldn't be necessary, but after 1-2 hours not crashing I would try it with powertop, too.

Update, now it has been 4 hours without crashes (ubuntu usb with powertop installed) without running powertop autotune.