TrueNAS SCALE on QNAP TS-653D - Heartbeat

[rtc]

Sharing my experience and hoping for some insight. I installed TrueNAS SCALE 22.02.0.1 on my QNAP TS-653D and unfortunately had to roll back to QNAP QTS 5.0.0.1986 Build 20220324 after an exhaustive effort to resolve the heartbeat.

My QNAP motherboard is Rev. 1, BIOS version is Q04MAR12, has 32GB of Crucial memory, a QM2-2S-220A PCIe with 2 each 2TB NVMe WDC model WDS200T2B0B-00YS70, and 4 each Seagate 6TB model ST6000VN001-2BB186 drives.

After installing SCALE on a USB 3.0 flash drive, I set up a single RAIDZ1 pool with the WDC drives and copied approximately 2.5 TB of data to the pool via SMB share. All worked great as expected, on Friday. However, On Saturday, I noticed a constant rhythmic sequential scan of the drives 1 through 4 approximately every 5 or 6 seconds. It was like a heartbeat and a bit loud, similar to the sound from the movie Contact (1997) with Matthew McConaughey. I thought it might have something to do with ZFS but after reviewing the logs and waiting several days for it to go away, I could not find anything that stood out.

I only reverted back because of the constant heartbeat cycle, not good background noise for online meetings and I am not sure what effect it could have on my drives.

Until I find a solution, I will continue to research this issue with later versions of TrueNAS SCALE in hopes of making a permanent transition. I am open to any ideas, and suggestions.

RTC

Show : my QNAP TS-653D

Model name: TS-653D Rev. 1
CPU: Intel(R) Celeron(R) J4125 CPU , up to 2700 MHz (4 cores, 4 threads)
BIOS version: Q04MAR12
Total memory: 32 GB (31 GB usable)
Memory slots: 2 (16 GB / 16 GB)

PCIe: QM2-2S-220A with 2 each 2TB NVMe WDC model WDS200T2B0B-00YS70
Drives: 4 each Seagate 6TB model ST6000VN001-2BB186 drives in RAIDZ1

 

[sretalla]

Until I find a solution,

You can consider this found...

Writes every 5 seconds are from collectd, which collects stats and stores them in the system dataset (which will have automatically moved itself to your first pool when you created it).

You can either move the sytem dataset back to the boot pool (knowing that the very reason for the automatic movement away from the boot pool was to avoid that load on USB sticks), or replace your boot media with some kind of SSD and then move it, or just add an SSD as an additioal pool and move it to that pool.

Problem solved (as documented in at least 10 other threads, I'm sure)

 

[rtc]

You can consider this found...

Writes every 5 seconds are from collectd, which collects stats and stores them in the system dataset (which will have automatically moved itself to your first pool when you created it).

You can either move the sytem dataset back to the boot pool (knowing that the very reason for the automatic movement away from the boot pool was to avoid that load on USB sticks), or replace your boot media with some kind of SSD and then move it, or just add an SSD as an additioal pool and move it to that pool.

Problem solved (as documented in at least 10 other threads, I'm sure)

sretalla, Thank you very much for your expeditious response. You are correct, more than a dozen responses were found to the inquiry "every 5 seconds", my bad for not digging deeper. Super excited about transitioning soon, will follow through on moving the sytem dataset to a new SSD.

Again, thank you for your help.