6.10.3 Parity Read Error (Self Caused?)

[Arbadacarba]

OK, So I've had this happen twice... And I feel like I was doing similar things both times... Namely starting a VM that I use for Cloning Unnasigned drives... I've been trying to improve the tools I use and I think it might be something I've done... But I don't see the mistake I'm making.

 

Recently on Jun 20 12:59:32 - My Parity drive suffered a read error and fell offline.

 

I had this happen once before in the last week or so and ran the extended smart tests with no failures... Further the Disk that failed is my newest disk but has been running without fail for several months.

 

I put it back in service and rebuilt the parity with no trouble but it just did it again today while I was trying to edit a couple of the VM's that I'm using (Clonzilla and GPartEd).

 

Thanks for any input. I'm ordering a new 18tb EXOS to install as parity, and I will RMA the 14TB

jupiter-diagnostics-20220620-1322.zip

[JorgeB]

Looks like a power/connection problem, suggest replacing/swapping cables/slot to rule that out.

[Arbadacarba]

It's just weird that it's all of a sudden an issue... But there's always a chance ive degraded some connection with the other things I've been doing...

 

I was just concerned that I was casuing it by connecting another /Dev/SD$ in a vm config.

[JorgeB]

1 minute ago, Arbadacarba said:

I was just concerned that I was casuing it by connecting another /Dev/SD$ in a vm config.

Don't see how that be be related, sometimes users pass-trough a controller by mistake to a VM, so when the VM starts all disks connected there are lost to Unraid, but that's not the case here.

[Arbadacarba]

Thanks for taking a look... I didn't think I was doing it wrong, but it seemed to coincide.

 

I suppose it's possible I'm getting vibration... There is an extension on that connection that I should possibly get rid of. Or at the very least check it for solid connection.

 

oh, and I have definitely passed through the wrong controller back when I was first setting things up.

[Arbadacarba]

Further input for interests sake: I mounted a drive using Unassigned Devices and instantly lost the parity drive to a read error...

 

The unassigned drive is inserted in a slot load tray above my parity drive, and not in the best of health. (Hence the reason I'm trying to access it.)

 

This slot load bay is also in use when I was launching the VM's that I thought were causing me the problem before...

 

I'm waiting on the replacement drive this week.

 

Oh the drive the system is repeatedly warning about is the drive I mounted to do data recovery on.

 

 

jupiter-diagnostics-20220621-0933.zip

[trurl]

20 hours ago, JorgeB said:

Looks like a power/connection problem, suggest replacing/swapping cables/slot to rule that out.

 

[trurl]

13 minutes ago, Arbadacarba said:

mounted a drive using Unassigned Devices and instantly lost the parity drive to a read error...

 

The unassigned drive is inserted in a slot load tray above my parity drive, and not in the best of health.

Did you actually insert this drive while the server was on?

[Arbadacarba]

Yes, it's on a hot swap capable Sata connection 

 

I've been using this bay as hot swap for over 4 years... Though the Sata controller has changed a few times over the years. The current connection is on the NZXT Mainboard and the port is set to Hot Plug in BIOS.

 

IOMMU group 5:[8086:43d2] 00:17.0 SATA controller: Intel Corporation Device 43d2 (rev 11)

[7:0:0:0] disk ATA TOSHIBA MQ01ABD1 2D /dev/sda 1.00TB

[8:0:0:0] disk ATA Samsung SSD 870 1B6Q /dev/sdc 2.00TB

[9:0:0:0] disk ATA Samsung SSD 870 1B6Q /dev/sdd 2.00TB

 

The Samsung Drives are my Cache Drives

 

My Array is entirely connected to a Dell Perc 310:

 

IOMMU group 18:[1000:0072] 06:00.0 Serial Attached SCSI controller: Broadcom / LSI SAS2008 PCI-Express Fusion-MPT SAS-2 [Falcon] (rev 03)

[3:0:0:0] disk ATA ST14000NE0008-2J EN01 /dev/sde 14.0TB

[3:0:1:0] disk ATA ST12000VN0007-2G SC60 /dev/sdf 12.0TB

[3:0:2:0] disk ATA ST10000NE0004-1Z EN01 /dev/sdg 10.0TB

[3:0:3:0] disk ATA ST10000VN0008-2P SC61 /dev/sdh 10.0TB

[3:0:4:0] disk ATA ST10000VN0008-2P SC61 /dev/sdi 10.0TB

[3:0:5:0] disk ATA ST10000VN0004-1Z SC60 /dev/sdj 10.0TB

 

The First Drive is my Parity... The only differentce with the Parity drive is that it is mounted to a seperate plate from the other Array disks, and there is a short Sata extension cable connected... AND IT"S ON THE SAME POWER LEAD AS THE REMOVABLE BAY!!!

 

(OK thought of that as I typed...)

 

[Arbadacarba]

The Power Supply is a beQuiet 1000W that is nowhere near the limits... Is there a way to see if the drive looses power?

[JorgeB]

Could just be causing a voltage drop on that cable.

[Arbadacarba]

The Power Supply is a beQuiet 1000W that is nowhere near the limits... Is there a way to see if the drive looses power?

 

I wonder if using a separate line for the Parity drive might help... I don't think it should, if I'm reading the manual correctly, but maybe just moving it a little farther away... Worth a shot.

 

Edited June 22, 2022 by Arbadacarba
Typos

[trurl]

1 hour ago, Arbadacarba said:

beQuiet 1000W that is nowhere near the limits

  

That says it has 4 12V rails. The power is split between those and disks only get 1 of the 4

[Arbadacarba]

That was my read as well... So even if I use a seperate cable it will not isolate the one drive from the external bays.

[JorgeB]

5 hours ago, Arbadacarba said:

The Power Supply is a beQuiet 1000W that is nowhere near the limits..

I didn't mean a PSU limit, but you could have a cable issue.

[Arbadacarba]

Considered wiring a Cap into the 12V rail at the plug for the Hotswap bay, but tried plugging a separate cable in for it rather than running it off the cable that I had in place... Seems to be holding now. Yay, time to start building the parity again.