Hi,

Recently I discovered the uptime of my server regularly reset because of unexpected reboots.

The last one occured 4 hours ago, but each time I don't see any clue in the syslog, example:

Jan 15 03:46:44 SylTower emhttpd: read SMART /dev/sde

Jan 15 03:47:02 SylTower kernel: mdcmd (54): set md_write_method 1

Jan 15 03:47:02 SylTower kernel:

Jan 15 04:01:56 SylTower emhttpd: spinning down /dev/sde

Jan 15 04:02:03 SylTower kernel: mdcmd (55): set md_write_method 0

Jan 15 04:02:03 SylTower kernel:

Jan 15 04:58:53 SylTower rc.rsyslogd: Syslog server daemon... Started.

Jan 15 04:58:53 SylTower emhttpd: Starting services...

Jan 15 04:58:53 SylTower sysDrivers: SysDrivers Build Complete

Jan 15 04:58:53 SylTower emhttpd: shcmd (72): /etc/rc.d/rc.samba reload

Starting 04:58:53 I think those are the logs when the server is starting, and before it's just runtime logs.

Previous unexpected reboots happened 13/01 10pm, 13/01 6am, 12/01 4:23pm, 11/01 7am.

Before today you'll see a lot of those errors:

nvme 0000:04:00.0: [ 0] RxErr (First)

pcieport 0000:00:1d.0: AER: Multiple Correctable error message received from 0000:04:00.0

nvme 0000:04:00.0: PCIe Bus Error: severity=Correctable, type=Physical Layer, (Receiver ID)

nvme 0000:04:00.0: device [15b7:5017] error status/mask=00000001/0000e000

I know it's because my NVME SSD is PCIe gen3 whereas the motherboard is PCIe gen4, I added "pci=noaer pcie_aspm=off" to the syslinux configuration to try to fix that, I don't think it's related to the crash.

I'm planning to downgrade Unraid to v6 because I think it began with one of the last OS update.

Any help much appreciated 🙏

syltower-diagnostics-20260115_1022.zip

Server rebooting by itself is almost always a hardware problem, start by running memtest.

Thanks @JorgeB I'll do memtest right now.

Ok first memtest pass is successful, I'll let it run a couple of other pass just to be sure.

Edit: 4 passes successful. I reset BIOS settings (I played with ASPM recently, that may be it), restarted Unraid and now running a CPU stress test (with "docker run --rm -it --cpus="4" polinux/stress stress --cpu 4 --timeout 3600")

Edited January 16Jan 16 by Syl20p

Hello,

Just to let you know, since last reboot with the syslinux configuration change, so far no reboot and no "PCIe Bus Error". Wanna wait some days before closing the post.

Hello, ok this morning my nas was completely down.

The power was on, but screen black and nothing respond.

I forced shutdown by pressing the power button, and from then the nas was unable to start, I coul'nt even access BIOS ! Only black screen.

So I made some tests and discovered the nas boot correctly when the usb stick (on which unraid is installed) is not plugged in.

I plugged in on another usb port and the server booted fine (but I'm at work today so cannot make further tests until this evening).

So it seems it's related either to the usb stick or the usb port of the motherboard (I hope it's the first!).

Meanwhile a bonus question : is this possible to install unraid on an SSD instead of usb stick?

2 hours ago, Syl20p said:

Meanwhile a bonus question : is this possible to install unraid on an SSD instead of usb stick?

Not yet, but it will be from 7.3

Hi, it's been more than week and the server never rebooted again, so I think it was related to the usb port of the motherboard.