Truenas core server keeps shutting down with no log entries

[sqwob]

working with half the ram in alternating slots didn't eliminate anything.

Disconnected all sata drives now (sata and power) and checking if it'll also power down.

Update: even without any pool hard drives present, it shutdown. removed the extra Intel NIC now (so it's just boot volume, mobo cpu & half the ram now)

 

[Davvo]

You sure the CPU cooler is working as intended?

 

[sqwob]

You sure the CPU cooler is working as intended?

yeah it's nearly cold to the touch

 

[Davvo]

yeah it's nearly cold to the touch

That's not proof of it working though: there might be contact issues between the CPU's HS and the cooler contact plate. Look at the temperature readings.

 

[WI_Hedgehog]

That's not proof of it working though: there might be contact issues between the CPU's HS and the cooler contact plate. Look at the temperature readings.

Don't be that crazy "it might be the CPU-paste" guy.

He only hit 38°C, should be fine unless he's getting hot-spots (or it's not the CPU that's overheating), but again, should be fine.

I'd guess it's a core component issue, CPU, RAM, any of the various RealTek chips (i've had bad luck with them when they get old--they're cheap, and for multiple reasons), plug-in board, some sort of acceleration setting that's pushing past the system limits now that the system is aging, could be anything. I'd unplug everything and boot off a USB diagnostic drive and see if it still trips over itself, and if not start adding things back in and re-testing.

I'll take back the CPU filter cap suggestion, the filter caps are all solid so it's probably not them. Could be the Power Supply Unit filter caps though, I bought an o-scope just to look at power--dirty, dirty man...no, just kidding, I run really clean PSUs and there's actually very little PSU noise, most of what I have is generated by cheap Chinese fans (Noctua fans run really clean, by the way, so you get what you pay for in that sense).

 

[sqwob]

Testing with different motherboard, ram en cpu now #sparepartsbin

To bad it doesn't fit in the case, otherwise this could be a definitive solution.

 

[sqwob]

Due to lack of spare parts and age of the hardware I've decided to purchase a new motherboard, ram and CPU

• Crucial Pro DDR4 32GB(2x16GB)
• AMD Ryzen 5 5500
• Asus Prime B550M-K, mATX, 4xDDR4
• LSI 9272-8i8-Port 6Gbps PCI-e Update 9207-8I flashed IT MODE

 

[sqwob]

All hardware finally arrived and have it back up and running. now for some testing!

• the Lsi 9272-8i Controller Card works like a charm (for sale on ebay preflashed in IT mode for 30€ including cables for 8 sata drives)
• The "chenyang CY M.2 NGFF Key B+M SATA to SATA 3.0 6Gbps 5 Ports Adapter Converter Port Multiplier Selector JMB575 2280" I tried as a temporary solution while waiting for the new controller card was a big disappointment (wouldn't even detect the attached drives). Never buy these.

 

[MrGuvernment]

Test away and let us know how it goes!

 

[chuck32]

Due to lack of spare parts and age of the hardware

That's the most frustrating part really. You can and want to trouble shoot yourself and quickly run out of compatible hardware for swaps. Naturally it makes more sense to upgrade the platform than buying spare parts for the existing platform.. I feel you.

Good luck with the new hardware!

 

[sqwob]

It's back as stable as it ever was :)

Glad to have this whole thing behind me and the setup modernised a little bit.

 

[WI_Hedgehog]

Are you going to get the old components in a test rig to find the source of the problem?

 

[sqwob]

Are you going to get the old components in a test rig to find the source of the problem?

Put them aside for now, maybe in the future, if I can find a compatible CPU somewhere to test with...