Scrub Error on all Pools

[andreaconfa]

Hello everybody,
my TrueNas has been giving me problems for months.

At each scrub that is run I receive errors (scrub repaired XX kb) on both my pools as follows:

- ZFS Pool consisting of 4 3TB disks CONNECTED TO CONTROLLER1 SATA PCIE
- RAID2 Pool consisting of two 4TB disks CONNECTED TO CONTROLLER2 SATA PCIE

The SATA controllers are separated for each pool so i assume that the issue is not here.

Even if i rase the errors, they recur again at the next scrub.
I have performed smart tests on all the disks but I don't know how to interpret them ... even if it seems strange that there are errors on both pools.
At the moment I am running a MEMTEST of the whole system to make sure that there are no errors at the ram level.

I attach all the smart tests.
Thank you

 

[Arwen]

Please use more precise terminology for your pools. I am guessing your "ZFS Pool" is a ZFS RAID-Z1 and the "RAID2" pool is a ZFS Mirror.

Further, it's helpful to describe all your hardware, and the exact TrueNAS version.

 

[andreaconfa]

Please use more precise terminology for your pools. I am guessing your "ZFS Pool" is a ZFS RAID-Z1 and the "RAID2" pool is a ZFS Mirror.

Further, it's helpful to describe all your hardware, and the exact TrueNAS version.

sorry for my poor english.
Yes you are guessing right.
Anyway i'm using the last version on TrueNAS Core running on a ESXi Hypervisor with the two sata controller directly attacched to the VM.
Previously, however, TrueNAS was installed directly on the hardware and the problem still existed.

The machine is an i5-4570 running on a B85-G43 Gaming.
16 GB of ram reserved for the TrueNAS VM

The two controller used is:
- https://www.amazon.it/gp/product/B07KF83M5W/ref=ppx_yo_dt_b_search_asin_title?ie=UTF8&psc=1
- https://www.amazon.it/gp/product/B07MBFWH81/ref=ppx_yo_dt_b_search_asin_title?ie=UTF8&psc=1

But again, the error is present on both pools connected to different controller

Running MemTest on the server machine i found that two of the four stick used is having some error... maybe is this the cause?

 

[Johnnie Black]

Running MemTest on the server machine i found that two of the four stick used is having some error... maybe is this the cause?

Likely yes.

 

[andreaconfa]

Ram is now ok but the issue still there

 

[Evertb1]

Running TrueNAS as a VM on ESXi is something that should be done only with great care and with proper hardware. None of your hardware is really recommended for the job. As far as I can see your controllers are based on a Marvell chipset. Those are not the most reliable.

 

[andreaconfa]

The same issue was present before the installation of esxi, when trueness was running standalone

 

[Arwen]

If your TrueNAS Core was having problems, introducing another layer of complexity, VMWare ESXi, would not be recommended.

I'd suggest fixing the RAM, (or out right leaving it off if you still have 8GBs after). Then restoring the TrueNAS server to bare metal, without VMWare. We can then look at fixing the problems with the pool.


But regardless, a gaming motherboard, desktop CPU and generic SATA cards are not necessarily the best for a server. This is not meant as a criticism, but to show that we, FreeNAS / TrueNAS users, tend to be conservative on our hardware. We do this because we want reliability over other things, (like re-using left over hardware).

ZFS will tend to find bad hard because it's designed that way. One of the original developers during ZFS' initial testing had problems with a PC he was testing on. He could not figure out why ZFS was reporting disk errors. Turns out he had a bad disk cable, (if I remember correctly, could have been a bad disk...). So, it turns out ZFS was doing exactly what it should, and he did not know it before with other file systems.

 

[andreaconfa]

I’m a home user, not a business so my only choise is a consumer motherboard.
I restore the standalone installation with working 8gb of ram now (tested 24h with memtest) keeping the two sata controller card
Then if someone can help me to find the problem I would appreciate it very much.

I replaced all the sata cable with new one arrived today from amazon

What more information I need to provide to help understand the problem?