Hi, Im tried of dealing with consistent issues, and this is beyond my knowledge to diagnose.

Long story short, I just had 2 drives disabled a few hours after rebuilding another drive, and one of these drives (Parity 2) was also replaced just recently (like 48h before, the other repair started), and I had run a preclear on it (default settings). Do note these are all used drives (I got them new, but where running for like 2 years on another project, with very limited read/write). So I'm not ruling them out due to previous usage. A month before, I also had similar issues to this, but didn't think much of it at the time. I am tried of dealing with issues every 1 - 3 months and looking at any suggestions to fix having to deal with this. I'm not ruling out just replacing everything, because at this point i just want the issues to stop.

I attached the diagnostics for the most recent incident.

nas-diagnostics-20260205-0918.zip

Edited February 5Feb 5 by Wurmatron
Likely issue found, considering solved.

SMART for parity2 and disk2 looks OK, though neither have had recent (or any) extended self-test completed.

Can't tell anything about filesystem without the array started in Normal mode.

Do that and post new diagnostics.

Here, can run the extended test if needed but that'l take a bit.

Note: This is after a restart.

nas-diagnostics-20260205-1040.zip

Emulated disk2 is mounted so should be OK to rebuild on top.

Do any other disks show SMART warning (👎) on the DASHBOARD page?

Looks like all disks except parity(sdb) are on this controller:

01:00.0 Serial Attached SCSI controller [0107]: Broadcom / LSI SAS2008 PCI-Express Fusion-MPT SAS-2 [Falcon] [1000:0072] (rev 03)
	Subsystem: Broadcom / LSI 9211-8i [1000:3020]
	Kernel driver in use: mpt3sas
	Kernel modules: mpt3sas

Could it be overheating?

no, all are green, except Parity 2, Disk 2, which are disabled.

Its possible, but I've got the internal case fan, Mini Desktop, and an external Fan pointing into the system, along with fans pointing at the array itself.

You think its a good idea to rebuild or should I run the extended test first.

When you had problems before, was it these same disks? That might indicate disk problems, or could indicate connection problems unless you changed cables or ports.

If different disks are involved, that might indicate controller. Or maybe power, but you don't have too many disks.

For the parity, yes, it was parity 2, though I ended up swapping drives around so its not the same connection anymore. It was originally on that PCI card, but not anymore. So all issues have been on the pci card, if thats what you are asking.

I just checked the temp of the pci card and its 65c so that does seem a bit high i think, ill try adding another fan, while im at it, once this backup is done. (my old one was 2 days ago, rather not lose data if i can help it.)

From those first diagnostics syslog

Jan 23 12:59:54 NAS kernel: sd 7:0:5:0: [sdh] tag#940 UNKNOWN(0x2003) Result: hostbyte=0x00 driverbyte=DRIVER_OK cmd_age=3s
Jan 23 12:59:54 NAS kernel: sd 7:0:5:0: [sdh] tag#940 Sense Key : 0x2 [current] 
Jan 23 12:59:54 NAS kernel: sd 7:0:5:0: [sdh] tag#940 ASC=0x4 ASCQ=0x0 
Jan 23 12:59:54 NAS kernel: sd 7:0:5:0: [sdh] tag#940 CDB: opcode=0x8a 8a 00 00 00 00 00 06 6c 4c 90 00 00 01 00 00 00
Jan 23 12:59:54 NAS kernel: I/O error, dev sdh, sector 107760784 op 0x1:(WRITE) flags 0x0 phys_seg 32 prio class 0
Jan 23 12:59:54 NAS kernel: md: disk0 write error, sector=107760720

disk0 write error disabled parity(1). Doesn't look like a disk problem though.

Then several days later

Feb  2 04:08:02 NAS kernel: sd 7:0:3:0: [sdf] tag#273 UNKNOWN(0x2003) Result: hostbyte=0x00 driverbyte=DRIVER_OK cmd_age=0s
Feb  2 04:08:02 NAS kernel: sd 7:0:3:0: [sdf] tag#273 Sense Key : 0x2 [current] 
Feb  2 04:08:02 NAS kernel: sd 7:0:3:0: [sdf] tag#273 ASC=0x4 ASCQ=0x0 
Feb  2 04:08:02 NAS kernel: sd 7:0:3:0: [sdf] tag#273 CDB: opcode=0x8a 8a 00 00 00 00 02 76 81 88 e0 00 00 00 20 00 00
Feb  2 04:08:02 NAS kernel: I/O error, dev sdf, sector 10578135264 op 0x1:(WRITE) flags 0x0 phys_seg 4 prio class 0
Feb  2 04:08:02 NAS kernel: md: disk3 write error, sector=10578135200

So not the same disks as now.

Did you change connections on parity(1) since then?

30 minutes ago, trurl said:

all disks except parity(sdb) are on this controller

It is currently on the motherboard controller.

Just now, trurl said:

Did you change connections on parity(1) since then?

No that wouldn't be right. The connections I was looking at earlier were from those same diagnostics.

So parity(1) was disabled while on the motherboard controller.

I swapped what was originally my Backup drive for parity 2, I think. (Didn't have an 18TB on hand when it died the first time)

How are these disks powered? Any splitters?

Ideally, no more than 4 disks per PSU cable. MOLEX-SATA splitters are better than SATA-SATA splitters.

All cables must have enough slack so they aren't pulling on the connector. Don't bundle data cables.

An external PSU to molex -> sata power, 3 per, for a total of 6 drives. (7 now,4 on one, now that i've got an another drive precleaning, just in case)

35 minutes ago, Wurmatron said:

Mini Desktop

2 minutes ago, Wurmatron said:

external PSU

I think I'm getting the picture now.

What can you tell me about that PSU?

Antec 350w, you think its the PSU? never even considered that. Note this PSU is to power the 7 drives. (Was 6, and still had issues)

Edited February 5Feb 5 by Wurmatron
More info

Not sure. Rebuild, parity check, turbo write (if you use that) runs all disks at the same time.

I have 7 HDD on my main server with 450W PSU

ok, its least you where able to give me a direction when I had none, others just constantly fighting issues.

May be worth me cheking the power requirements of each drive and comparing to what the PSU says it can output, and see what I have leftover.

I also keep an eye on the card's temp, mabey its just overheating, causing odd issues.