Version 6.9.2 - unclean shutdown

[Potter117]

new to unraid, and this is my first post, so take it easy on me lol.  Please forgive any weird typos since I am doing this from my phone and autocorrect is funny sometimes.  
 

Recently I moved from a power edge server to a custom built computer because I needed a quieter setup in my home office.  On the new computer, it runs a parity check for about 90 minutes, then stays up for maybe half an hour.  Every time I check on it, the array is down and it says “Unclean shutdown”.  I have searched the forums but have not found anything that helps me.  I am not great at this kind of stuff, so please forgive me if I am making an obvious mistake.  My board is a mortar b450m titanium from MSI.  I have an RX560 graphics card installed.  Processor is Ryzen 5 1600.  Ram is Corsair ram rated for 3200mhz.  When I moved from the old hardware to the new hardware, I shut down the server, moved the USB to the new setup, booted, and added new drives because I couldn’t use the sas drives from the power edge.  One drive is a 1TB ssd and one is a WD Gold 1TB drive.  I do have a backup of everything (small 1TB Plex server that hasn’t grown much is what this is for), so I can try anything to get this going. 

I have included a screenshot of my dashboard and the diagnostic zip that I downloaded while typing this.  I set the priority to urgent, since this does make the system unusable.  Please let me know if there is another priority I should be using for this type of issue.  Thank you in advance for any advice.  

wall-server-diagnostics-20210608-2144.zip

[trurl]

I copied your "bug report" to a new topic in General Support. Usually you should post a new topic in General Support to see if we can help you with your problem, and if it seems to be an actual bug then we can go on to reporting it.

 

Why are you using nvme as parity?? You should use that as cache and put another spinner as parity.

 

For Ryzen crashes, see here:

 

Also, syslog resets on reboot, so setup Syslog Server so syslog can be retrieved after crash:

 

[Potter117]

I am using an NVME for parity because unraid sees it as larger, and the ONLY two drives I have that aren't sas drives are the nvme and the WD Gold.  Sooooo.......  That's why.
I'm rolling with for now lol.  I have switched from Windows to Mac now that I am out of school and don't need the versatility.  I may put a virtual windows machine on the unraid server once it's up and running again.  I followed those 2 links.  I disabled the xmp ram profiles and disabled the global c-state control.  If it is still running, uninterrupted, this time tomorrow we are in business. 

Thanks for the guidance.

Edited June 9, 2021 by Potter117
fixing typo

[Potter117]

So far it has been up for 13 hours and some change.  Yesterday it wouldn't get 30 minutes past the completed parity check.  Once it hits 24 I'll take another look and we will see what it's doing.  If it's still going strong at that point we can probably close this.

[trurl]

13 hours ago, Potter117 said:

NVME for parity because unraid sees it as larger

Parity does have to be at least as large as the largest data disk in the array, but I was suggesting using another spinner as parity. Seems a waste to use nvme as parity since parity won't even be accessed when reading your data, and your dockers and VMs would perform better if setup on SSD pool (cache) outside the parity array. There are other reasons to not mix SSDs and HDDs in the array 

 

[Potter117]

I understand what you meant, but I literally have no other drives in hand besides these two.  Even tho they are both listed as 1TB, unraid is seeing the NVME as larger.  Once I have this sorted and a larger disk for my parity, I’ll make better use of the nvme.