Rebooting and I have no idea why

TrueNAS General
1

I recently built a little machine with an asus X99 Deluxe Mobo, Xeon E5-2698 v4, 128gb sk hynix ECC ram, a 1080ti and 1 2tb m.2, 2 1tb sata SSDs, and 2 2tbHDDs.

I asked AI to suggest the best pool arrangement and I have the main drives as the HDDs in mirror (data vdev), the 2 1tb SSDs and the metadata storage, and the 2tb m.2 as the Cache.

I keep getting random reboots and when I turn the monitor on for the system i get something that looks like some sort of memory error. I did memtest86 for over 40 hours and 0 failures. I have no idea what the issue is, I am a total noob. I tried to upload some screen shots but it says you cant?

Anyways, the monitor keeps rapidly going through errors. one of them reads like this:

  1. 2357.791484] EDAC MCO: 43 CE memory read rain:32 syndrome: 0x0 *: 1 bank-group: 0) 2357. 794699] error on CPU_SrcID#0_Ha#0_Chan#1_DIMM#0 (channel:1 slot:0

  2. 2357.7957651 - OVERFLOW area:DRAM err_code: 0001:0091 socket:0 ha:0 channel_mask:2 rank: row:oxfffe col: RAS: Soft-offlining pfn: 0x203fba2 Memory 2357. 796823] failure: buffer EDAC 2357.797877] EDAC bridge MC0: HANDLING MCE MEMORY ERROR overflow when queuing memory

If anyone has any idea what I am doing wrong (I havent done much of anything yet) please let me know.

2
  1. You used AI - and it fed you what is almost certainly shit suggestions. I cannot state catagorically that they are shit as I don’t know your use case.
  2. You have a memory error, you need to determine if the error is with a DIMM or a memory slot

The memory error is the most important issue. First remove all the memory, then put back 50% and see what happens - work from there

3

The error is telling you there is an issue with the module of channel 1 slot 0, which is probably the first stick in the second channel: remove both modules of that channel and see what happens.

According to the motherboard’s manual, it should be slot B1.

I suggest checking your kernel log.

1 Like
4

Looks like a slot was bad. I was able to source another identical mobo and now that error is gone! Thank you. In your opinion, what would the optimal set up for the drive be?

5

Thanks man, slot was no good. got a replacement mobo and all is well. Do you see a problem with the way my drives are set up? I dont really know what the best set up is. I dont really have anything important on there. I really dont want to use this as a NAS for my other systems, I want this machine to be a stand alone machine that I just use to run apps and maybe some VMs. totally for fun.

6

I doubt you need L2ARC with 128 GB of RAM and your usage. Mirroring the HDs for slower and larger data storage and mirroring the SSDs for aps and VMs is probably the best choice. What are you using for the boot drive?

BASICS

iX Systems pool layout whitepaper

Special VDEV (sVDEV) Planning, Sizing, and Considerations

1 Like
7

Maybe TN is overkill for that, and Linux distro XYZ could be more straightforward… your call.

If you don’t require large amount of storage, ditch the HDDs and run a simple 2-way mirror VDEV of NVMe SSDs. Doing so would allow you to ditch the L2ARC and metadata VDEV as well.

The only real issue with your current setup is that the “cache” drive is too large for your RAM size… you want it to be around 4 to 6 times your RAM due to performance reasons.

1 Like
8

some crappy 128gb ssd I had lying around. Maybe I should get another M.2 and mirror that also.

9

the SSD for boot should be fine. No need to mirror for simple use. You just keep a current copy of your configuration file and secret seeds and a copy of the install iso.

You can replace the boot device with a new one, do a clean install and then upload your config file and you should be back to running.

1 Like