I'm at my wits end with this Server

[matthewprime83]

Hey everyone,

 

Thanks for reading but I'm at my wits end with this home server that I have been trying to build. After weeks of trouble shooting and trying to get it to be stable it just wont work. After turning on the server it will work for about 48 hours before shutting off and throwing a million errors at me saying everything from disks are dead to the disks not having file structure to hardware issues. After this last shut down my Disk 1 says that it is unmontable: No File system even though it was working just a few hours ago with most of my files on it. I'm using all new components and will attach a list of what I have and all the Smart-Log errors that I can find to upload. Any help that ya'll would be able to give me would be amazing... been trying to use this unraid server to help my family use Plex during this pandemic. Thank you again for reading, SGT Housel.

 

Parts List:

 

Mother Board: ASRock AB350M Pro4

CPU: AMD Ryzen 5 1600 Six-Core @3200 Mhz

Memory: 16 GIB DDR4 3200Mhz

GPU: GTX 1060- 6 Gigabyte 

Hard Drives:

Palarity- Seagate IronWolf 10 TB

Disk 1- Seagate IronWolf 8 TB

Disk 2- Seagate Barracuda 4TB

Disk 3- Western Digital Blue 4TB

Disk 4- Western Digital Blue 4TB

Disk 5- Seagate Barracuda Compute 8TB

Disk 6- SSD 512GB

Disk 7- SSD 512GB

 

Cache- SSD 512GB

             SSD 512GB

 

 

Disk 1- tower-smart-20200406-1052.zip Tower_FixProblems.pdf Tower_Dashboard.pdf Tower_Syslog.pdf

[Radium_Angel]

16 minutes ago, matthewprime83 said:

unraid

I''m not an expert on unraid, but isn't mixing and matching drives a bad idea?

[matthewprime83]

I do not know... I haven't heard anything about it, but if its true then it will take ALOT longer to get this up and running since money is tight right now.

 

[2FA]

13 minutes ago, Radium_Angel said:

I''m not an expert on unraid, but isn't mixing and matching drives a bad idea?

It's not, there isn't any sort of striping. Parity is stored on a specific disk which is occasionally checked.

 

37 minutes ago, matthewprime83 said:

 

Use a different Linux live USB to see if the disks mount fine with something other than Unraid.

[2FA]

Also, that IronWolf has a lot of seek errors and ECC errors so try to recover data off of it, then try to get it RMA'd.

[matthewprime83]

Ok, I will give that a try! I went into the BIOS and changed the timing on the Ram from 2400 to 1866 to see if that will provide some stability for first generation Ryzen. The Unraid Forum said that could help with the shutdowns that keep happening every 72 hours.

[Dutch_Master]

Regular failures sound like a timing issue to me. Not sure where or how to solve it, but you may have a misconfigured system here. Recover your data ASAP then either re-install Unraid or try a plain Linux distro and build from that.

[Jarsky]

Apr 6 06:16:46 Tower kernel: mce: [Hardware Error]: Machine check events logged
Apr 6 06:16:46 Tower kernel: mce: [Hardware Error]: CPU 7: Machine Check: 0 Bank 5: bea0000000000108
Apr 6 06:16:46 Tower kernel: mce: [Hardware Error]: TSC 0 ADDR 1ffff8168b28a MISC d012000101000000 SYND 4d000000 IPID 500b000000000
Apr 6 06:16:46 Tower kernel: mce: [Hardware Error]: PROCESSOR 2:800f11 TIME 1586168176 SOCKET 0 APIC 9 microcode 8001129

 

Make sure you're running the correct BIOS version on your motherboard. 

According to the documentation here: http://www.asrock.com/mb/AMD/AB350M Pro4/#BIOS

With your Ryzen 5 1600 (Summitridge) you should be on BIOS 5.90. If you're below this you need to upgrade to 5.50 first, and then upgrade to 5.90. If you're above this you should downgrade as 6.00+ is not supported for Summitridge. 

 

Memory: 16 GIB DDR4 3200Mhz

3200Mhz memory is not supported for 1st gen Ryzen. Drop the clocks to 2666Mhz on the memory. 

 

 

 

As for disk SMART values, these are the ones i'd keep an eye on

ID# ATTRIBUTE_NAME          FLAGS    VALUE WORST THRESH FAIL RAW_VALUE
  1 Raw_Read_Error_Rate     POSR--   079   064   044    -    75577771
  7 Seek_Error_Rate         POSR--   082   060   045    -    172750270
195 Hardware_ECC_Recovered  -O-RC-   006   002   000    -    75577771

The values look OK as long as they aren't increasing at a drastic rate. 

 

 

The problem though is your crashing, as UnRAID array really dislikes losing power. Check your BIOS to ensure correct Microcode, and drop the memory frequency. That should stabilize your system. 

[Salacious B Crum]

1st gen ryzen will run at 3200 but it was very particular about which brands and they didn't play nice with SK hynix my 1600x and 1800x are both running samsung B-die G-Skill Flare X @3200 pricey even then but worked right outta the box if your overclocking that could pose a problem as mentioned make sure you are using a usb 2 stick and are using it in a usb 2 port that has caused many issues with unRAID although I have been running a usb3 drive in a usb2 port for over 3 years now with no issues