X10SL7-F cpu temp during rsync

proto

Patron
Joined
Sep 28, 2015
Messages
269
I think I have a problem with my Supermicro X10SL7-F motherboard that I mounted on an old aluminium tower case by moving it from a Node 804.
The CPU temperature rises to 98 ° C and the motherboard alarm siren starts playing shortly after launching an rsync task to synchronize a dataset between two pools.

The CPU fan is that stock of the Xeon E3-1230 processor.
When copying with rsync the values reported by the IPMI are these:

Code:
CPU Temp         | 98.000     | degrees C  | cr    | 0.000     | 0.000     | 0.000     | 95.000    | 98.000    | 100.000   
System Temp      | 32.000     | degrees C  | ok    | -9.000    | -7.000    | -5.000    | 80.000    | 85.000    | 90.000   
Peripheral Temp  | 38.000     | degrees C  | ok    | -9.000    | -7.000    | -5.000    | 80.000    | 85.000    | 90.000   
PCH Temp         | 50.000     | degrees C  | ok    | -11.000   | -8.000    | -5.000    | 90.000    | 95.000    | 100.000   
[...]
FAN1             | 2000.000   | RPM        | ok    | 400.000   | 600.000   | 800.000   | 25300.000 | 25400.000 | 25500.000
FAN2             | 1700.000   | RPM        | ok    | 100.000   | 100.000   | 100.000   | 25300.000 | 25400.000 | 25500.000
FAN3             | 1100.000   | RPM        | ok    | 100.000   | 100.000   | 100.000   | 25300.000 | 25400.000 | 25500.000
FAN4             | 1400.000   | RPM        | ok    | 100.000   | 100.000   | 100.000   | 25300.000 | 25400.000 | 25500.000
FANA             | 200.000    | RPM        | nc    | 100.000   | 100.000   | 200.000   | 25300.000 | 25400.000 | 25500.000
[...]


FAN1 is the CPU fan.
The other fans (Noctua) for the time being I have arranged with minimum values.

Currently FreeNAS is virtualized on ESXi 6.7, while the disks are all in passthru (LSI on mb + Dell HBA).

Should I worry about my build? :smile:
 

Evertb1

Guru
Joined
May 31, 2016
Messages
700
Rsync is CPU intensive so some heath in the CPU is to be expected. But with temps that high you are around the max I think. And if your HBA is very active it will also add to the heath. Improving the airflow in your case and adding a good quality fan for your sas controller would not be a bad thing. If you have a good airflow the boxed cpu cooler should be sufficient but buying a better one would not do any harm.
 

proto

Patron
Joined
Sep 28, 2015
Messages
269
I agree with you about the heaviness of the rsync task, but before updating the system I could safely transfer data between a system with X10SL7-F and a HP Microserver n36l provided that I do not perform compression during the transfer via network.
Of course I also have problems with ventilation right now and I will probably have to go back and put the X10 back inside the Node 804 or buy a new case.

Here is my failing HW build:

IMG_1657.png


This is an old Lian-Li pc-a10... It was a good case, but I suspect I have a broken airflow :smile:
 

DrKK

FreeNAS Generalissimo
Joined
Oct 15, 2013
Messages
3,630
Sir,

There is no universe, to my thinking, where this configuration should go to 98C, even under maximum load. You MUST have a problem with either the CPU fan, *or* the heatsink mounting/heatsink compound. It's literally impossible.
 

Apollo

Wizard
Joined
Jun 13, 2013
Messages
1,458
Not impossible.
I believe I have seen this kind of temps on my Xeon 1231 V2 as well.
 

proto

Patron
Joined
Sep 28, 2015
Messages
269
It's strange. I just detached and re-attached all the hardware from one box to another, without changing anything. Well, apart from hard disks and adding an HBA.

Now I remember one thing that happened last week after changing the case. The siren began playing on the first boot in the UEFI shell and then towards the end of the BIOS update.

The other day I installed FreeNAS as vm inside ESXi 6.7, fixed the fan values, made some tests and everything seemed to work until I started to transfer the data both with rsync and via zfs send / receive. I can still use cp, at 80 degrees ...

However, I bought a bigger sink (noctua) and tomorrow I should be able to mount it and see if it dissipates better.
 

DrKK

FreeNAS Generalissimo
Joined
Oct 15, 2013
Messages
3,630
Guys, if you're seeing 98C on your Xeon 1231, something is wrong. I don't really want to argue about it. You don't have to believe me if you don't want to.
 

proto

Patron
Joined
Sep 28, 2015
Messages
269
I hope I have solved! Thank you all.

This morning I replaced the CPU heatsink (intel stock) with a Noctua tower cooler. Now I am transferring 1.1 TB of data (mostly photo / video) from a half hour with a stable average temperature of about 36C.

I just have to fight with the fans again :smile: But this is another topic.
 

DrKK

FreeNAS Generalissimo
Joined
Oct 15, 2013
Messages
3,630
Indeed.

Glad you got it solved sir.
 
Top