HPL test failed Fester recommended burn-in test (screenshot)

[naspcboss]

I found this thread to break in the NAS. The Breakin stress test (standard test) has been running for 3.5 hrs and there are 2 HPL test fails. Is this something to concern about? Are there better ways to stress test the system (since ECC is not tested on this software)?

Thanks in advance!

 

[Chris Moore]

Those failures appear to be related to the temperature. I suggest that you need a better cooling solution to keep the that under control.

 

[naspcboss]

This is using the stock Intel cooler. Do you mean I need to get a better cooler? The chassis is the Supermicro Super Chassis.

Maybe Breakin is too old for newer systems (last update 2015). What alternatives do you use? I also need to test ECC of the RAM.

 

[Chris Moore]

This is using the stock Intel cooler. Do you mean I need to get a better cooler?

With at temperature that high, I would say it might need something a bit better. There must not be good airflow through the case. What model chassis is it?

I just do the hard drive burn-in and count on the server board and ECC memory to just work.

 

[Mihalich]

It's time :)

 

[Chris Moore]

It's time :)

That is pretty funny.

 

[naspcboss]

Thanks for the replies!

1. Chassis is Superchassis CSE-721TQ-250B 250W. Idle CPU temps are around 39C. Is that too high?
2. Adding a cooler wont help with the PCH being high though. Is that temp an issue?
I'm not going to transcode on this. I do plan to use a VM to sync data to a cloud provider which I dont think will cause a lot of CPU usage.

3. Do you use this guide for HDD burn-in or something else?

Right now BMC reports with nothing running (no VMs)
CPU Temp 38 degrees C
PCH Temp 45 degrees C

 

[Chris Moore]

CSE-721TQ-250B

That is a similar chassis to the one the FreeNAS folks use for the FreeNAS Mini. It is known to run hot and several posts have been made about upgrading the fans and adding additional fans to the chassis to prevent overheating.
I don't know if you will be able to find the specific posts but I recall seeing several over the past few years. The main rear fan needs an upgrade to give more overall airflow to keep the drive temps down and a fan blowing across the system board under the drive cage to get more airflow across the PCH. It is just a bad chassis for airflow. It is not a bad chassis if you are using a super low power Intel® Atom processor or something like that, but a full power Xeon system generates much more heat.

 

[Mihalich]

Idle CPU temps are around 39C. Is that too high?

The temperature in the server room is about 18 degrees.

 

[Chris Moore]

Idle CPU temps are around 39C. Is that too high?

I wouldn't say that the idle temp is too high because it is not outside the manufacturers range of acceptable temperature, however I know from what other users of that chassis type have stated in their posts, the drives also end up being too hot over time if there is no upgrade to the fan configuration.