Hardware or Software RAID fails drives test perfect.

[unknownerrors]

Hello
I’m having a great deal of difficulty with setting up a working riad configuration. I am setting up 3 arrays. 1 made of 5 SATA3 6TB drives, and 1 made of 5 10tb sata3 drives and lastly 1 made up of 5 6TB SAS drives.

all 3 of these arrays fail under a few configurations. ideally all three would be in use at the same time but I have also tested having only one at a time connected thinking maybe its a power issue.

I have also connected all of these drives (with the exception of the SAS drives, they must be connection to a raid controller.) connected directly to my mobo and did a number of health tests that all pass with flying colors on the drives are brand new so I wouldn’t expect anything less.

To test I used H.D.S. Hungary Hard Disk Sentinel PRO v6.30.
using brand new 4 port SATA break away cables (when connected to the RAID Controller.

I have tested the raid arrays with two controller types both fail they are
LSI MegaRAID SAS 9260-4i (ir mode firmware) connected to a Intel RES2SV240 expander
and
a lsi megaraid sas 3 3008 fury (it mode firmware) also coonnected to the expander listed above.

its worth noting that I have 2 of the expanders both end up with the same problem.

I have used windows versions Server edition 2012 RC storage server,
2019 Server, and 2025 Server.

I need the iSCSI target server, have have ZERO desire to use TrueNAS because I know little to nothing about the underlaying OS and in the event I need to utilize recovery tools (such as R-Studio) i’d have to take the drives out of the TrueNAS system and place them in a Windows box (what a pain up in the neck!)

When I create the array’s either in disk management or in the controller bios every thing goes smoothly. I format the array in either REFS or NTFS still everything goes fine. Its not until I begin writing data the the array that things go sideways. if it was configured in the bios the volume just drops, poof gone until I restart the system. if it was created in disk management (software raid) when I start writing data to it also drops the volume but the disks go into a failed state in disk management, not missing, just failed.

So it seems to me that when writing data to multiple disk concurrently the system just can’t handle it and drops the volume. If I use these disk one at a time (no t raided) they work great. With that in mind I’m leaning toward a power issue but that confuses me even more because the power supply i’m using puts out 850W that should be more than enough to run 15 drives let alone 5. well 6 when you count the OS SSD…

If I had hair i’d be pulling it out right now…
Here’s a link to my server systems Dxdiag.txt file if it helps

[type or paste code here](https://www.mediafire.com/file_premium/mpyx6zli9f6y5s8/DxDiag.txt/file)

If there is anything I can provide that would be help please just ask.

[diizzy]

Try another PSU and what do the LSI cards say? I’d also drop the expander card for the sake of troubleshooting. Are you also using some kind of backplane or are they directly connect using cables?
Given that you have trouble troubleshooting Windows I dont really see the point in saying that anything else is out of the window. That being said, you’d likely find it easier to diagnose if you run either FreeBSD or some Linux distro which keeps ZFS and the kernel fairly up to date and by default they’ll be very verbose.

[unknownerrors]

When you say

what do the LSI cards say? Blockquote
Are you asking what the card say on them physically, what do they say in the boot process, in windows? please clarify your question.

I forgot to mention that I did try another PSU., same results. I also tried a new mobo, same results.

Right now I have the 5 6TB SATA drives connected directly to the mobo and the 5 6TH SAS drives connected to the 9260-4i controller (IR mode). I am trying to move a large amount of data from the SATA drive (software raid 5) to the SAS drive (hardware RAID0) The SAS volume keeps dropping from the system under high load and does not return until I reboot.
I have tested these connected directly to the controller with two 9260-4i cards with the same result.

does it matter that I have the cards slotted in PCI-E slots marked for SLI/Crossfire?

[diizzy]

Again, logs…

Does it also occur on the 3008 (when connected directly), same cabling used?

[unknownerrors]

How you like system logs?
From event viewer? MSM (for the hardware raid)?
I do not know enough about FreeBSD or Linux to use them for this task.

[unknownerrors]

Also, no. all drive connect direct to either odd-in card or direct to the MB. no backplane in use.

[wendell]

have seen this kinda thing when the hba/adapter is running too hot, or borderline.

[unknownerrors]

That could very well be the issue. the drives typically run around 70 - 80*C getting even hotter under load.
I’m using

Configured in the same manner as the image above shows, the only difference is that I have removed the front lockable panel to increase air flow.

This makes prefect sense. The drives work as intended when not configured in some sort of raid because the heat generated by one drive is a good deal less than what is generated by 5x3 …

I’ll look into better cooling options and get back to you with the results .