Intel Socket 1151 Motherboards with IPMI AND Support for iGPU


Recommended Posts

Hi Guys,

 

I need some recommendations before buying my first Unraid server in order to replace some old synology devices.

 

Im planning to buy a ASRock Rack E3C246D4U2-2L2T with an Intel Xeon E-2288G. Does anybody can confirm if that motherboard have the same issues with iGPU and IPMI as the ASRock Rack E3C246D4U2? Or is everything fine with this motherboard?

 

I already contacted the ASRock Support in order to get some informations about known issues with iGPU and IMPI also because the motherbaord have still its first BIOS, but the ASRock Support just answered some crap.

 

Thanks in advance,

d1ck13

Link to comment
30 minutes ago, d1ck13 said:

Hi Guys,

 

I need some recommendations before buying my first Unraid server in order to replace some old synology devices.

 

Im planning to buy a ASRock Rack E3C246D4U2-2L2T with an Intel Xeon E-2288G. Does anybody can confirm if that motherboard have the same issues with iGPU and IPMI as the ASRock Rack E3C246D4U2? Or is everything fine with this motherboard?

 

I already contacted the ASRock Support in order to get some informations about known issues with iGPU and IMPI also because the motherbaord have still its first BIOS, but the ASRock Support just answered some crap.

 

Thanks in advance,

d1ck13

 

This newer board looks great!  I like the extra network adapter and hdmi.

 

I just have the E3C246D4U2.

 

Who did you talk with at ASRock?

 

Maybe you can entice @Hoopster to upgrade his board and let you know if it works?  LOL.

  • Like 1
Link to comment
5 hours ago, d1ck13 said:

...

I already contacted the ASRock Support in order to get some informations about known issues with iGPU and IMPI also because the motherbaord have still its first BIOS, but the ASRock Support just answered some crap.

 

Thanks in advance,

d1ck13

I'm in Germany and spoke with the same folks the people here have spoken with. Super helpful.

 

Just know that they may not be able to answer at Asrock because this may just be an issue with Unraid. In my tests which I posted about a few pages back the IPMI works just fine at the same time the iGPU is active in Proxmox.

Edited by Plaidy
Link to comment
24 minutes ago, parisv said:

Not sure if I should undo this an enable the gpu driver in the first link?

I went ahead and removed the items from my "go" file after performing the "touch" command to enable the driver. All working fine. Anecdotally, hardware transcoding seems to be using more CPU than I remember, but I think I could easily just not be remembering correctly 😂

Link to comment
1 hour ago, kaiguy said:

I went ahead and removed the items from my "go" file after performing the "touch" command to enable the driver. All working fine. Anecdotally, hardware transcoding seems to be using more CPU than I remember, but I think I could easily just not be remembering correctly 😂

 

It should be using the gpu though?

Link to comment

After not encountering it since pretty much the first few weeks of building this new server, I got the IPMI motherboard temp warning tonight. I didn't realize this was still a thing. Running the latest L2.34 BIOS. Are others still getting this from time to time?

Link to comment

Well, at 4:11am this morning my server had its first hard lockup. I am not home but VPN'd in via my router and saw that IPMI logged a Run-time Critical Stop. Nothing has changed with my hardware setup, so I have to assume something has changed with 6.9.0 that has exposed some instabilities with my system. So strange. Hopefully no one else starts experiencing issues.

Link to comment
On 3/5/2021 at 12:50 AM, kaiguy said:

Are others still getting this from time to time?

Since updating to 6.9.0 and the accompanying reboot 5 1/2 days ago, the unRAID WebGUI via the System Temp plugin has been reporting proper MB temps in the 26-31 C range, but IPMI continues to report MB temps from 83-88C. 

 

Who knows how long it will be until the System Temp plugin decides to once again join the madness. 

 

No lockups though so there is that.

Link to comment
3 hours ago, Hoopster said:

No lockups though so there is that.

Are you still running with Turbo Boost disabled, @Hoopster? I never had to disable it before, but I wonder if I'm not being plagued with that system lock.

 

Running a

grep MHz /proc/cpuinfo

while the system is running a parity check from the system lock results in

cpu MHz		: 3500.338
cpu MHz		: 3500.505
cpu MHz		: 3500.004
cpu MHz		: 3500.169
cpu MHz		: 3500.220
cpu MHz		: 3500.430
cpu MHz		: 3500.135
cpu MHz		: 3499.397
cpu MHz		: 3499.443
cpu MHz		: 3500.380
cpu MHz		: 3500.302
cpu MHz		: 3502.510
cpu MHz		: 3500.185
cpu MHz		: 3500.332
cpu MHz		: 3500.376
cpu MHz		: 3500.164

but as I recall it was showing similar frequencies when idle... not sure if its normal for it to be at such a high freq all the time.

 

Hoping for no more locks, but I'm not confident.

Link to comment
45 minutes ago, kaiguy said:

Are you still running with Turbo Boost disabled

No, I enabled Turbo Boost a few months ago.

 

Through the Tips and Tweaks Plugin I had the CPU Scaling Governor set to Power Save and on this setting, the CPU would throttle down when idle; sometimes as low as 800 MHz.

 

I decided to install a Windows 10 VM yesterday and read that there were many performance issues with VMs with the Scaling Governor in Power Save mode.  Performance mode is recommended (or On Demand if CPU uses cpufreq driver rather that Pstate). 

 

I now have it on Performance with Turbo Boost enabled and I see very high CPU frequencies.

image.png.5831ad528d8f08c76203832a44de73d4.png

 

Here is Power Save with Turbo Boost disabled.

image.png.44edbdba51289f83fc7543a9dbac6b7c.png

 

And here it is after sitting idle for about 20 minutes with Power Save and Turbo Boost enabled, but I suspect I caught it before Turbo Boost kicked back in after being enabled.

image.png.c7570bfff51edef6a97faeb7b9535e2a.png

 

In all cases, the VM is running.

 

Edited by Hoopster
  • Thanks 1
Link to comment

@Hoopster I have a few VMs but by default they aren't running. I have had my system on Power Save and Turbo Boost since I started using this mobo.

 

I also think the lock happened during the mover process, and the only change is I started using an nvme for my cache instead of one of my SSDs when I switched to 6.9.0, so in retrospect perhaps that's the culprit.

 

In Performance, the CPU is at max freq the whole time, right? Think that's a material increase in power consumption? Heat?

Link to comment
3 minutes ago, kaiguy said:

Think that's a material increase in power consumption? Heat?

Not that I have seen.  The CPU temps look roughly the same in Performance mode as they did in Power Save mode.  I'm not sure Power Save really buys you much other than it taking longer to ramp up CPU speeds.

 

I suppose power consumption might depend on if you were doing any heavy in either unRAID or a VM.  At idle/low usage, I see no difference.

Link to comment

Hey @Hoopster I was curious since you've setup a VM are you passing any USB controllers through on the E3C246D4U? I'm trying to learn this currently and I think I'm coming to the conclusion that there is only one controller on the board. This is an issue since the unraid usb key is on this controller. Am I mistaken here or are am I understanding pass-through wrong.2070886958_ScreenShot2021-03-07at6_08_31PM.thumb.png.693f4694a0b09f649434fd699aff760e.png

Link to comment
12 hours ago, JimmyGerms said:
I'm coming to the conclusion that there is only one controller on the board. This is an issue since the unraid usb key is on this controller. Am I mistaken here or are am I understanding pass-through wrong.

You are correct.  The IOMMU groups on this board do not separate out the USB controllers as there appears to be only one.  If you want to pass through a USB controller, you will need to install a PCIe USB controller that should then be in a separate group.
 
I even installed a USB 2.0 card in the server connected to a motherboard USB 2.0 header for making sure I was booting unRAID from USB 2.0.  However, that is not a separate controller as all USB ports from the PCH are tied to the same controller.

Of course you could also try the acs override method to try and breakout a usb hub from the controller.

Edited by Hoopster
Link to comment

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.