Critical temperature detected suggest system shutdown

[ryanvox]

I am having an issue where I am getting the following error on my monitor:

coretemp6: critical temperature detected, suggest system shutdown

But when i go into the UI there are no alerts or anything indicating a critical temp issue. my CPU (i7-8700) is stable at about 40 Celsius. During a large load it will shoot up to maybe 70 Celsius for a second and the very highest but right now when i see those errors on the screen my CPU is only under like 5% load. Could this be a bug or something to be concerned about?

I did recently upgrade to TrueNas Core 13.0-U5.3. I had a SSD failure. I reflashed the newest version. This was about 5 days ago which is about when i noticed this issue....but on the flip side i never had a monitor plugged into it before the failure. so if there was no alerts on the UI i would never have known there was an issue.

 

[HoneyBadger]

During a large load it will shoot up to maybe 70 Celsius for a second and the very highest but right now when i see those errors on the screen my CPU is only under like 5% load. Could this be a bug or something to be concerned about?

Hey @ryanvox

Can you run the following command from a shell or SSH session to see if the returned values differ from your dashboard?

sysctl -a | egrep -E "cpu\.[0-9]+\.temp"

If your 70C shows up as a very sudden, rapid increase in temperatures, it's also possible that you do not have good contact between your CPU and heatsink. Consider checking the contact area and replacing the thermal interface material.

 

[ryanvox]

Hey @ryanvox

Can you run the following command from a shell or SSH session to see if the returned values differ from your dashboard?

sysctl -a | egrep -E "cpu\.[0-9]+\.temp"

If your 70C shows up as a very sudden, rapid increase in temperatures, it's also possible that you do not have good contact between your CPU and heatsink. Consider checking the contact area and replacing the thermal interface material.

The 70C shows up when i am transcoding multiple streams on Plex. I did the command and it returned the following

root@freenas:~ # sysctl -a | egrep -E "cpu\.[0-9]+\.temp"
dev.cpu.11.temperature: 41.0C
dev.cpu.9.temperature: 45.0C
dev.cpu.7.temperature: 41.0C
dev.cpu.5.temperature: 41.0C
dev.cpu.3.temperature: 41.0C
dev.cpu.1.temperature: 45.0C
dev.cpu.10.temperature: 40.0C
dev.cpu.8.temperature: 42.0C
dev.cpu.6.temperature: 41.0C
dev.cpu.4.temperature: 42.0C
dev.cpu.2.temperature: 42.0C
dev.cpu.0.temperature: 42.0C

If I do a sysctl -a | grep temper I get:

coretemp2: critical temperature detected, suggest system shutdown
coretemp6: critical temperature detected, suggest system shutdown

I saw that command online. Not sure really what it points to. I do have 2 other boards on my PCIE slots. I have a sata expansion card and SFP adapter card i use for a fiber connection between this truenas and my back up using Rsync for backups. Nothing is running though that i can see so not sure what the critical temp is from. I did look into the server. All fans are spinning.

 

[HoneyBadger]

Those temperatures look normal for an 8700, and if the 70C is a "plateau" of heat rather than a "spike" then it's likely not a hardware/TIM issue.

coretemp would indicate that it's alerting on a hot CPU - it might just be that a sensor is being misread somehow, as TjMAX on that CPU is 100C, so you shouldn't see an alert from a core or two at 70C.

 

[ryanvox]

Those temperatures look normal for an 8700, and if the 70C is a "plateau" of heat rather than a "spike" then it's likely not a hardware/TIM issue.

coretemp would indicate that it's alerting on a hot CPU - it might just be that a sensor is being misread somehow, as TjMAX on that CPU is 100C, so you shouldn't see an alert from a core or two at 70C.

Ok thank you. i'll try repasting and i'll probably throw on a better cooler. The current one is just a stock cooler.